{ "generated": "2026-10-21", "models": [ { "id": "whisper-large-v3", "name": "Whisper large-v3", "org": "OpenAI", "type": "STT", "released": { "value": "2023-11-07", "source": "https://huggingface.co/api/models/openai/whisper-large-v3", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "1.54B", "source": "https://huggingface.co/api/models/openai/whisper-large-v3", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/openai/whisper-large-v3" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/openai/whisper-large-v3", "note": "HF card: Apache-2.0; GitHub repo: code and weights MIT" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su" ], "source": "https://huggingface.co/openai/whisper-large-v3" }, "streaming": { "value": false, "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "30-second window encoder-decoder; real-time use needs a wrapper such as whisper_streaming" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~10 GB VRAM", "gb": 10, "kind": "VRAM", "source": "https://github.com/openai/whisper#available-models-and-languages" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/openai/whisper-large-v3" }, "links": { "huggingface": "https://huggingface.co/openai/whisper-large-v3", "github": "https://github.com/openai/whisper" }, "hf_downloads_30d": 3755126, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "whisper-large-v3-turbo", "name": "Whisper large-v3-turbo", "org": "OpenAI", "type": "STT", "released": { "value": "2024-10-01", "source": "https://huggingface.co/api/models/openai/whisper-large-v3-turbo", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "809M", "source": "https://huggingface.co/api/models/openai/whisper-large-v3-turbo", "note": "safetensors parameter count" }, "license": { "value": "mit", "source": "https://huggingface.co/openai/whisper-large-v3-turbo" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/openai/whisper-large-v3-turbo", "note": "MIT" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su" ], "source": "https://huggingface.co/openai/whisper-large-v3-turbo" }, "streaming": { "value": false, "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "30-second window model; streaming only via wrappers" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~6 GB VRAM", "gb": 6, "kind": "VRAM", "source": "https://github.com/openai/whisper#available-models-and-languages" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/openai/whisper-large-v3-turbo" }, "links": { "huggingface": "https://huggingface.co/openai/whisper-large-v3-turbo", "github": "https://github.com/openai/whisper" }, "hf_downloads_30d": 6094388, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "whisper-tiny", "name": "Whisper tiny", "org": "OpenAI", "type": "STT", "released": { "value": "2022-09-26", "source": "https://huggingface.co/api/models/openai/whisper-tiny", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "38M", "source": "https://huggingface.co/api/models/openai/whisper-tiny", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/openai/whisper-tiny" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/openai/whisper-tiny", "note": "Apache-2.0 on HF card; MIT on GitHub" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su" ], "source": "https://huggingface.co/openai/whisper-tiny" }, "streaming": { "value": false, "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "30-second window model; streaming only via wrappers" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~1 GB VRAM", "gb": 1, "kind": "VRAM", "source": "https://github.com/openai/whisper#available-models-and-languages" }, "runs_on_cpu": { "value": true, "source": "https://github.com/SYSTRAN/faster-whisper#small-model-on-cpu", "note": "Whisper models run on CPU via faster-whisper int8" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/openai/whisper-tiny" }, "links": { "huggingface": "https://huggingface.co/openai/whisper-tiny", "github": "https://github.com/openai/whisper" }, "hf_downloads_30d": 818386, "benchmarks": [ { "task": "stt", "runner": "faster-whisper tiny (int8, beam 5)", "device": "cpu", "latency_first_segment_s": 0.294, "rtf": 0.0291, "audio_seconds": 10.09, "load_s": 0.46, "peak_rss_mb": 290.4, "wer": 0.0968, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open source speech models are improving every week. The benchmark measures how fast a model runs on every day hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:24+02:00", "device_id": "x86-cpu", "runtime": "faster-whisper", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-tiny.json" }, { "task": "stt", "runner": "whisper.cpp tiny (ggml fp16, beam 5, 8 threads, pywhispercpp)", "device": "cpu", "latency_first_segment_s": 0.461, "rtf": 0.0457, "audio_seconds": 10.09, "load_s": 0.08, "peak_rss_mb": 226.4, "wer": 0.0968, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open source speech models are improving every week. The benchmark measures how fast a model runs on every day hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:28+02:00", "device_id": "x86-cpu", "runtime": "whisper.cpp", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-tiny__whisper.cpp.json" } ], "benchmark_status": "measured" }, { "id": "whisper-base", "name": "Whisper base", "org": "OpenAI", "type": "STT", "released": { "value": "2022-09-26", "source": "https://huggingface.co/api/models/openai/whisper-base", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "73M", "source": "https://huggingface.co/api/models/openai/whisper-base", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/openai/whisper-base" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/openai/whisper-base", "note": "Apache-2.0 on HF card; MIT on GitHub" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su" ], "source": "https://huggingface.co/openai/whisper-base" }, "streaming": { "value": false, "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "30-second window model; streaming only via wrappers" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~1 GB VRAM", "gb": 1, "kind": "VRAM", "source": "https://github.com/openai/whisper#available-models-and-languages" }, "runs_on_cpu": { "value": true, "source": "https://github.com/SYSTRAN/faster-whisper#small-model-on-cpu", "note": "Whisper models run on CPU via faster-whisper int8 or whisper.cpp; measured here" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/openai/whisper-base" }, "links": { "huggingface": "https://huggingface.co/openai/whisper-base", "github": "https://github.com/openai/whisper" }, "hf_downloads_30d": 1652817, "benchmarks": [ { "task": "stt", "runner": "faster-whisper base (int8, beam 5)", "device": "cpu", "latency_first_segment_s": 0.413, "rtf": 0.041, "audio_seconds": 10.09, "load_s": 3.64, "peak_rss_mb": 443.0, "wer": 0.0645, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open Source Speech Models are improving every week. The Spenchmark measures how fast a model runs on everyday hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:35+02:00", "device_id": "x86-cpu", "runtime": "faster-whisper", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-base.json" }, { "task": "stt", "runner": "whisper.cpp base (ggml fp16, beam 5, 8 threads, pywhispercpp)", "device": "cpu", "latency_first_segment_s": 0.793, "rtf": 0.0786, "audio_seconds": 10.09, "load_s": 0.26, "peak_rss_mb": 334.7, "wer": 0.0323, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open Source speech models are improving every week. The benchmark measures how fast a model runs on everyday hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:40+02:00", "device_id": "x86-cpu", "runtime": "whisper.cpp", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-base__whisper.cpp.json" } ], "benchmark_status": "measured" }, { "id": "whisper-small", "name": "Whisper small", "org": "OpenAI", "type": "STT", "released": { "value": "2022-09-26", "source": "https://huggingface.co/api/models/openai/whisper-small", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "242M", "source": "https://huggingface.co/api/models/openai/whisper-small", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/openai/whisper-small" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/openai/whisper-small", "note": "Apache-2.0 on HF card; MIT on GitHub" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su" ], "source": "https://huggingface.co/openai/whisper-small" }, "streaming": { "value": false, "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "30-second window model; streaming only via wrappers" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~2 GB VRAM", "gb": 2, "kind": "VRAM", "source": "https://github.com/openai/whisper#available-models-and-languages" }, "runs_on_cpu": { "value": true, "source": "https://github.com/SYSTRAN/faster-whisper#benchmark", "note": "faster-whisper README benchmarks the small model on CPU (Intel Core i7-12700K, 8 threads)" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/openai/whisper-small" }, "links": { "huggingface": "https://huggingface.co/openai/whisper-small", "github": "https://github.com/openai/whisper" }, "hf_downloads_30d": 3610565, "benchmarks": [ { "task": "stt", "runner": "faster-whisper small (int8, beam 5)", "device": "cpu", "latency_first_segment_s": 1.017, "rtf": 0.1008, "audio_seconds": 10.09, "load_s": 4.11, "peak_rss_mb": 989.8, "wer": 0.0, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open source speech models are improving every week, this benchmark measures how fast a model runs on everyday hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:52+02:00", "device_id": "x86-cpu", "runtime": "faster-whisper", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-small.json" }, { "task": "stt", "runner": "whisper.cpp small (ggml fp16, beam 5, 8 threads, pywhispercpp)", "device": "cpu", "latency_first_segment_s": 2.984, "rtf": 0.2958, "audio_seconds": 10.09, "load_s": 0.78, "peak_rss_mb": 800.9, "wer": 0.0, "wer_note": "Clean synthetic Piper speech (en_US-lessac-medium), not a real-world WER", "transcript": "Open source speech models are improving every week, this benchmark measures how fast a model runs on everyday hardware, how much memory it needs, and whether it is practical to deploy.", "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:28:11+02:00", "device_id": "x86-cpu", "runtime": "whisper.cpp", "results_file": "bench/results/20261021-cloud-x86-8vcpu/whisper-small__whisper.cpp.json" } ], "benchmark_status": "measured" }, { "id": "faster-whisper-large-v3", "name": "faster-whisper (large-v3, CTranslate2)", "org": "SYSTRAN", "type": "STT", "released": { "value": "2023-11-23", "source": "https://huggingface.co/api/models/Systran/faster-whisper-large-v3", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "1.55B", "source": "https://github.com/openai/whisper#available-models-and-languages", "note": "same weights as Whisper large-v3, converted to CTranslate2" }, "license": { "value": "mit", "source": "https://huggingface.co/Systran/faster-whisper-large-v3" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/Systran/faster-whisper-large-v3", "note": "MIT" }, "languages": { "value": [ "en", "zh", "de", "es", "ru", "ko", "fr", "ja", "pt", "tr", "pl", "ca", "nl", "ar", "sv", "it", "id", "hi", "fi", "vi", "he", "uk", "el", "ms", "cs", "ro", "da", "hu", "ta", "no", "th", "ur", "hr", "bg", "lt", "la", "mi", "ml", "cy", "sk", "te", "fa", "lv", "bn", "sr", "az", "sl", "kn", "et", "mk", "br", "eu", "is", "hy", "ne", "mn", "bs", "kk", "sq", "sw", "gl", "mr", "pa", "si", "km", "sn", "yo", "so", "af", "oc", "ka", "be", "tg", "sd", "gu", "am", "yi", "lo", "uz", "fo", "ht", "ps", "tk", "nn", "mt", "sa", "lb", "my", "bo", "tl", "mg", "as", "tt", "haw", "ln", "ha", "ba", "jw", "su", "yue" ], "source": "https://huggingface.co/Systran/faster-whisper-large-v3" }, "streaming": { "value": false, "source": "https://github.com/SYSTRAN/faster-whisper#benchmark", "note": "Not built-in; README lists third-party streaming projects (Whisper-Streaming, WhisperLive)" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "~2.9 GB VRAM (large-v2, int8, beam 5, project benchmark)", "gb": 2.9, "kind": "VRAM", "source": "https://github.com/SYSTRAN/faster-whisper#benchmark", "note": "Project benchmark used large-v2, not v3; fp16 used ~4.5 GB" }, "runs_on_cpu": { "value": true, "source": "https://github.com/SYSTRAN/faster-whisper#benchmark", "note": "int8 CPU inference supported" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/Systran/faster-whisper-large-v3" }, "links": { "huggingface": "https://huggingface.co/Systran/faster-whisper-large-v3", "github": "https://github.com/SYSTRAN/faster-whisper" }, "hf_downloads_30d": 1164877, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "distil-large-v3.5", "name": "Distil-Whisper large-v3.5", "org": "Hugging Face", "type": "STT", "released": { "value": "2024-12-05", "source": "https://huggingface.co/api/models/distil-whisper/distil-large-v3.5", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "756M", "source": "https://huggingface.co/api/models/distil-whisper/distil-large-v3.5", "note": "safetensors parameter count" }, "license": { "value": "mit", "source": "https://huggingface.co/distil-whisper/distil-large-v3.5" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/distil-whisper/distil-large-v3.5", "note": "MIT" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/distil-whisper/distil-large-v3.5" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/distil-whisper/distil-large-v3.5" }, "links": { "huggingface": "https://huggingface.co/distil-whisper/distil-large-v3.5", "github": "https://github.com/huggingface/distil-whisper" }, "hf_downloads_30d": 8850, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "parakeet-tdt-0.6b-v3", "name": "Parakeet TDT 0.6B v3", "org": "NVIDIA", "type": "STT", "released": { "value": "2025-08-04", "source": "https://huggingface.co/api/models/nvidia/parakeet-tdt-0.6b-v3", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "600M", "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3", "note": "This model has 600 million model parameters" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3", "note": "CC-BY-4.0 (attribution); card: ready for commercial/non-commercial use" }, "languages": { "value": [ "en", "es", "fr", "de", "bg", "hr", "cs", "da", "nl", "et", "fi", "el", "hu", "it", "lv", "lt", "mt", "pl", "pt", "ro", "sk", "sl", "sv", "ru", "uk" ], "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3" }, "streaming": { "value": true, "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3", "note": "Chunked streaming via NeMo script" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "At least 2 GB RAM to load", "gb": 2, "kind": "RAM", "source": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/nvidia/parakeet-tdt-0.6b-v3" }, "links": { "huggingface": "https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3", "github": "https://github.com/NVIDIA/NeMo" }, "hf_downloads_30d": 626738, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "parakeet-unified-en-0.6b", "name": "Parakeet Unified EN 0.6B", "org": "NVIDIA", "type": "STT", "released": { "value": "2026-04-07", "source": "https://huggingface.co/api/models/nvidia/parakeet-unified-en-0.6b", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "600M", "source": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b" }, "license": { "value": "NVIDIA Open Model License", "source": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b", "note": "Card: ready for commercial/non-commercial use (NVIDIA Open Model License terms apply)" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b" }, "streaming": { "value": true, "source": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b", "note": "Offline + streaming in one model; latency 160 ms to 2080 ms" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/nvidia/parakeet-unified-en-0.6b" }, "links": { "huggingface": "https://huggingface.co/nvidia/parakeet-unified-en-0.6b", "github": "https://github.com/NVIDIA/NeMo" }, "hf_downloads_30d": 4447, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "canary-1b-v2", "name": "Canary 1B v2", "org": "NVIDIA", "type": "STT", "released": { "value": "2025-08-04", "source": "https://huggingface.co/api/models/nvidia/canary-1b-v2", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "978M", "source": "https://huggingface.co/nvidia/canary-1b-v2" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/nvidia/canary-1b-v2" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/nvidia/canary-1b-v2", "note": "CC-BY-4.0; card: ready for commercial/non-commercial use" }, "languages": { "value": [ "bg", "hr", "cs", "da", "nl", "en", "et", "fi", "fr", "de", "el", "hu", "it", "lv", "lt", "mt", "pl", "pt", "ro", "sk", "sl", "es", "sv", "ru", "uk" ], "source": "https://huggingface.co/nvidia/canary-1b-v2" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "At least 6 GB RAM to load", "gb": 6, "kind": "RAM", "source": "https://huggingface.co/nvidia/canary-1b-v2" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/nvidia/canary-1b-v2" }, "links": { "huggingface": "https://huggingface.co/nvidia/canary-1b-v2", "github": "https://github.com/NVIDIA/NeMo" }, "hf_downloads_30d": 172429, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "canary-qwen-2.5b", "name": "Canary-Qwen 2.5B", "org": "NVIDIA", "type": "STT", "released": { "value": "2025-06-26", "source": "https://huggingface.co/api/models/nvidia/canary-qwen-2.5b", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "2.56B", "source": "https://huggingface.co/api/models/nvidia/canary-qwen-2.5b", "note": "safetensors parameter count" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/nvidia/canary-qwen-2.5b" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/nvidia/canary-qwen-2.5b", "note": "CC-BY-4.0; card: ready for commercial use" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/nvidia/canary-qwen-2.5b" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/nvidia/canary-qwen-2.5b" }, "links": { "huggingface": "https://huggingface.co/nvidia/canary-qwen-2.5b", "github": "https://github.com/NVIDIA/NeMo" }, "hf_downloads_30d": 39670, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "moonshine-streaming-medium", "name": "Moonshine Streaming Medium", "org": "Moonshine AI (Useful Sensors)", "type": "STT", "released": { "value": "2026-01-06", "source": "https://huggingface.co/api/models/moonshine-ai/moonshine-streaming-medium", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "245M", "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium", "note": "Tiny 34M / Small 123M variants also available" }, "license": { "value": "mit", "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium" }, "commercial_use": { "value": "yes", "source": "https://github.com/moonshine-ai/moonshine#license", "note": "MIT (streaming models are MIT in all languages)" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium" }, "streaming": { "value": true, "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "Targets sub-1 GB memory edge devices", "gb": 1, "kind": "RAM", "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium", "note": "Design target, not a measured minimum" }, "runs_on_cpu": { "value": true, "source": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium", "note": "Built for edge-class hardware" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/moonshine-ai/moonshine-streaming-medium" }, "links": { "huggingface": "https://huggingface.co/moonshine-ai/moonshine-streaming-medium", "github": "https://github.com/moonshine-ai/moonshine" }, "hf_downloads_30d": 42339, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "moonshine-base", "name": "Moonshine Base", "org": "Moonshine AI (Useful Sensors)", "type": "STT", "released": { "value": "2024-11-02", "source": "https://huggingface.co/api/models/moonshine-ai/moonshine-base", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "61M", "source": "https://huggingface.co/moonshine-ai/moonshine-base" }, "license": { "value": "mit", "source": "https://huggingface.co/moonshine-ai/moonshine-base" }, "commercial_use": { "value": "yes", "source": "https://github.com/moonshine-ai/moonshine#license", "note": "MIT for English; legacy non-English non-streaming models are non-commercial" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/moonshine-ai/moonshine-base" }, "streaming": { "value": false, "source": "https://github.com/moonshine-ai/moonshine#license", "note": "Legacy non-streaming model" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://huggingface.co/moonshine-ai/moonshine-base", "note": "Designed for memory/compute-constrained platforms" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/moonshine-ai/moonshine-base" }, "links": { "huggingface": "https://huggingface.co/moonshine-ai/moonshine-base", "github": "https://github.com/moonshine-ai/moonshine" }, "hf_downloads_30d": 24242, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "kyutai-stt-2.6b-en", "name": "Kyutai STT 2.6B (EN)", "org": "Kyutai", "type": "STT", "released": { "value": "2025-06-06", "source": "https://huggingface.co/api/models/kyutai/stt-2.6b-en", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "2.62B", "source": "https://huggingface.co/api/models/kyutai/stt-2.6b-en", "note": "safetensors parameter count" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/kyutai/stt-2.6b-en" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/kyutai/stt-2.6b-en", "note": "CC-BY-4.0 (attribution)" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/kyutai/stt-2.6b-en" }, "streaming": { "value": true, "source": "https://huggingface.co/kyutai/stt-2.6b-en", "note": "2.5 s delay; 1B en/fr sibling has 0.5 s delay" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/kyutai/stt-2.6b-en" }, "links": { "huggingface": "https://huggingface.co/kyutai/stt-2.6b-en", "github": "https://github.com/kyutai-labs/delayed-streams-modeling" }, "hf_downloads_30d": 0, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "voxtral-mini-4b-realtime", "name": "Voxtral Mini 4B Realtime (2602)", "org": "Mistral AI", "type": "STT", "released": { "value": "2026-01-21", "source": "https://huggingface.co/api/models/mistralai/Voxtral-Mini-4B-Realtime-2602", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "4.43B", "source": "https://huggingface.co/api/models/mistralai/Voxtral-Mini-4B-Realtime-2602", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602", "note": "Apache-2.0, card states research and commercial use" }, "languages": { "value": [ "en", "fr", "es", "de", "ru", "zh", "ja", "it", "pt", "nl", "ar", "hi", "ko" ], "source": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602" }, "streaming": { "value": true, "source": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602", "note": "Natively streaming, causal audio encoder" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": "Single GPU with >= 16 GB memory (BF16, vLLM)", "gb": 16, "kind": "VRAM", "source": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/mistralai/Voxtral-Mini-4B-Realtime-2602" }, "links": { "huggingface": "https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602", "github": null }, "hf_downloads_30d": 2171152, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "qwen3-asr-1.7b", "name": "Qwen3-ASR 1.7B", "org": "Alibaba Qwen", "type": "STT", "released": { "value": "2026-01-28", "source": "https://huggingface.co/api/models/Qwen/Qwen3-ASR-1.7B", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "2.35B", "source": "https://huggingface.co/api/models/Qwen/Qwen3-ASR-1.7B", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/Qwen/Qwen3-ASR-1.7B" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/Qwen/Qwen3-ASR-1.7B", "note": "Apache-2.0" }, "languages": { "value": "30 languages + 22 Chinese dialects (52 total)", "source": "https://huggingface.co/Qwen/Qwen3-ASR-1.7B" }, "streaming": { "value": true, "source": "https://huggingface.co/Qwen/Qwen3-ASR-1.7B", "note": "Streaming only with the vLLM backend" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/Qwen/Qwen3-ASR-1.7B" }, "links": { "huggingface": "https://huggingface.co/Qwen/Qwen3-ASR-1.7B", "github": "https://github.com/QwenLM/Qwen3-ASR" }, "hf_downloads_30d": 1186346, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "vibevoice-asr", "name": "VibeVoice-ASR", "org": "Microsoft", "type": "STT", "released": { "value": "2026-01-21", "source": "https://huggingface.co/api/models/microsoft/VibeVoice-ASR", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "8.67B", "source": "https://huggingface.co/api/models/microsoft/VibeVoice-ASR", "note": "safetensors parameter count" }, "license": { "value": "mit", "source": "https://huggingface.co/microsoft/VibeVoice-ASR" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/microsoft/VibeVoice-ASR", "note": "MIT; Microsoft's repo advises further testing before commercial/real-world use" }, "languages": { "value": "50+ languages", "source": "https://github.com/microsoft/VibeVoice" }, "streaming": { "value": false, "source": "https://github.com/microsoft/VibeVoice", "note": "This checkpoint is offline; a separate VibeVoice-ASR-Streaming model exists" }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/microsoft/VibeVoice-ASR" }, "links": { "huggingface": "https://huggingface.co/microsoft/VibeVoice-ASR", "github": "https://github.com/microsoft/VibeVoice" }, "hf_downloads_30d": 686190, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "granite-speech-3.3-8b", "name": "Granite Speech 3.3 8B", "org": "IBM", "type": "STT", "released": { "value": "2025-04-14", "source": "https://huggingface.co/api/models/ibm-granite/granite-speech-3.3-8b", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "8.65B", "source": "https://huggingface.co/api/models/ibm-granite/granite-speech-3.3-8b", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/ibm-granite/granite-speech-3.3-8b" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/ibm-granite/granite-speech-3.3-8b", "note": "Apache-2.0" }, "languages": { "value": [ "multilingual", "en", "fr", "de", "es", "pt" ], "source": "https://huggingface.co/ibm-granite/granite-speech-3.3-8b" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": "n/a", "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/ibm-granite/granite-speech-3.3-8b" }, "links": { "huggingface": "https://huggingface.co/ibm-granite/granite-speech-3.3-8b", "github": "https://github.com/ibm-granite/granite-speech-models" }, "hf_downloads_30d": 8809, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "kokoro-82m", "name": "Kokoro 82M", "org": "hexgrad", "type": "TTS", "released": { "value": "2024-12-26", "source": "https://huggingface.co/api/models/hexgrad/Kokoro-82M", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "82M", "source": "https://huggingface.co/hexgrad/Kokoro-82M" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/hexgrad/Kokoro-82M" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/hexgrad/Kokoro-82M", "note": "Apache-2.0; card welcomes production/commercial deployment" }, "languages": { "value": [ "en-US", "en-GB", "es", "fr", "hi", "it", "ja", "pt-BR", "zh" ], "source": "https://github.com/hexgrad/kokoro", "note": "HF metadata lists only en; lang codes from the kokoro package README" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": null, "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://github.com/thewh1teagle/kokoro-onnx", "note": "ONNX build (kokoro-onnx) runs on CPU; measured here on an 8 vCPU x86 VM" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/hexgrad/Kokoro-82M" }, "links": { "huggingface": "https://huggingface.co/hexgrad/Kokoro-82M", "github": "https://github.com/hexgrad/kokoro" }, "hf_downloads_30d": 10533404, "benchmarks": [ { "task": "tts", "runner": "kokoro-onnx fp32 (onnxruntime CPU) voice=af_heart", "device": "cpu", "time_to_first_audio_s": 1.416, "rtf": 0.1234, "audio_seconds": 11.48, "load_s": 1.11, "peak_rss_mb": 933.9, "runs": 5, "notes": "create_stream(): first audio = first sentence chunk. Peak RSS includes model load.", "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:11+02:00", "device_id": "x86-cpu", "runtime": "kokoro-onnx", "results_file": "bench/results/20261021-cloud-x86-8vcpu/kokoro-82m.json" }, { "task": "tts", "runner": "kokoro-onnx int8 (onnxruntime CPU) voice=af_heart", "device": "cpu", "time_to_first_audio_s": 1.686, "rtf": 0.148, "audio_seconds": 11.39, "load_s": 0.64, "peak_rss_mb": 744.0, "runs": 5, "notes": "create_stream(): first audio = first sentence chunk. Peak RSS includes model load.", "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:21+02:00", "device_id": "x86-cpu", "runtime": "kokoro-onnx", "results_file": "bench/results/20261021-cloud-x86-8vcpu/kokoro-82m__int8.json" } ], "benchmark_status": "measured" }, { "id": "f5-tts", "name": "F5-TTS", "org": "SWivid (SJTU et al.)", "type": "TTS", "released": { "value": "2024-10-07", "source": "https://huggingface.co/api/models/SWivid/F5-TTS", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": null, "source": null }, "license": { "value": "cc-by-nc-4.0", "source": "https://huggingface.co/SWivid/F5-TTS" }, "commercial_use": { "value": "no", "source": "https://github.com/SWivid/F5-TTS#license", "note": "Code MIT, pretrained weights CC-BY-NC-4.0 (Emilia training data)" }, "languages": { "value": null, "source": null }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://github.com/SWivid/F5-TTS", "note": "Zero-shot from reference audio + ref_text" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/SWivid/F5-TTS" }, "links": { "huggingface": "https://huggingface.co/SWivid/F5-TTS", "github": "https://github.com/SWivid/F5-TTS" }, "hf_downloads_30d": 853842, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "xtts-v2", "name": "XTTS v2", "org": "Coqui", "type": "TTS", "released": { "value": "2023-10-31", "source": "https://huggingface.co/api/models/coqui/XTTS-v2", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": null, "source": null }, "license": { "value": "Coqui Public Model License (CPML)", "source": "https://huggingface.co/coqui/XTTS-v2" }, "commercial_use": { "value": "no", "source": "https://coqui.ai/cpml", "note": "CPML is a non-commercial license" }, "languages": { "value": "17 languages", "source": "https://huggingface.co/coqui/XTTS-v2" }, "streaming": { "value": true, "source": "https://github.com/idiap/coqui-ai-TTS", "note": "README: XTTS can stream with <200 ms latency" }, "voice_cloning": { "value": true, "source": "https://huggingface.co/coqui/XTTS-v2", "note": "6-second reference clip, cross-language" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/coqui/XTTS-v2" }, "links": { "huggingface": "https://huggingface.co/coqui/XTTS-v2", "github": "https://github.com/idiap/coqui-ai-TTS" }, "hf_downloads_30d": 6302234, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "piper", "name": "Piper (voices)", "org": "Rhasspy / Open Home Foundation", "type": "TTS", "released": { "value": "2023-06-22", "source": "https://huggingface.co/api/models/rhasspy/piper-voices", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": null, "source": null }, "license": { "value": "Voices repo: MIT; engine (piper1-gpl): GPL-3.0", "source": "https://huggingface.co/rhasspy/piper-voices" }, "commercial_use": { "value": "conditional", "source": "https://github.com/OHF-Voice/piper1-gpl", "note": "Engine is GPL-3.0; each voice's training data has its own license, check the voice's MODEL_CARD" }, "languages": { "value": [ "ar", "ca", "cs", "cy", "da", "de", "el", "en", "es", "fa", "fi", "fr", "hu", "is", "it", "ka", "kk", "lb", "lv", "ne", "nl", "no", "pl", "pt", "ro", "ru", "sk", "sl", "sr", "sv", "sw", "tr", "uk", "vi", "zh" ], "source": "https://huggingface.co/rhasspy/piper-voices" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": null, "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://github.com/OHF-Voice/piper1-gpl", "note": "Fast local neural TTS, ONNX on CPU (smoke-tested here)" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/rhasspy/piper-voices" }, "links": { "huggingface": "https://huggingface.co/rhasspy/piper-voices", "github": "https://github.com/OHF-Voice/piper1-gpl" }, "hf_downloads_30d": 0, "benchmarks": [ { "task": "tts", "runner": "piper-tts (onnxruntime) voice=en_US-lessac-medium", "device": "cpu", "time_to_first_audio_s": 0.051, "rtf": 0.0174, "audio_seconds": 9.95, "load_s": 1.44, "peak_rss_mb": 378.9, "runs": 5, "hardware_label": "Cloud x86 VM, 8 vCPU Intel Xeon, 16 GB RAM, no GPU", "hardware": { "cpu": "Intel(R) Xeon(R) Processor", "cpu_threads": 8, "ram_gb": 15.6, "gpu": null, "os": "Linux 6.12.94+", "python": "3.13.5" }, "date": "2026-10-21T23:27:02+02:00", "device_id": "x86-cpu", "runtime": "piper-tts", "results_file": "bench/results/20261021-cloud-x86-8vcpu/piper.json" } ], "benchmark_status": "measured" }, { "id": "orpheus-3b", "name": "Orpheus 3B (0.1 ft)", "org": "Canopy Labs", "type": "TTS", "released": { "value": "2025-03-17", "source": "https://huggingface.co/api/models/canopylabs/orpheus-3b-0.1-ft", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "3.78B", "source": "https://huggingface.co/api/models/canopylabs/orpheus-3b-0.1-ft", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/canopylabs/orpheus-3b-0.1-ft" }, "commercial_use": { "value": "yes", "source": "https://github.com/canopyai/Orpheus-TTS", "note": "Apache-2.0 (HF repo is gated: accept terms to download)" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/canopylabs/orpheus-3b-0.1-ft" }, "streaming": { "value": true, "source": "https://github.com/canopyai/Orpheus-TTS", "note": "~200 ms streaming latency, ~100 ms with input streaming" }, "voice_cloning": { "value": true, "source": "https://github.com/canopyai/Orpheus-TTS", "note": "Zero-shot cloning; README notes it was not explicitly trained for it" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": true, "source": "https://huggingface.co/api/models/canopylabs/orpheus-3b-0.1-ft" }, "links": { "huggingface": "https://huggingface.co/canopylabs/orpheus-3b-0.1-ft", "github": "https://github.com/canopyai/Orpheus-TTS" }, "hf_downloads_30d": 17093, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "sesame-csm-1b", "name": "Sesame CSM 1B", "org": "Sesame", "type": "TTS", "released": { "value": "2025-03-06", "source": "https://huggingface.co/api/models/sesame/csm-1b", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "1.55B", "source": "https://huggingface.co/api/models/sesame/csm-1b", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/sesame/csm-1b" }, "commercial_use": { "value": "yes", "source": "https://github.com/SesameAILabs/csm", "note": "Apache-2.0 (HF repo gated)" }, "languages": { "value": [ "en" ], "source": "https://github.com/SesameAILabs/csm", "note": "Some non-English capacity but 'likely won't do well'" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": null, "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": true, "source": "https://huggingface.co/api/models/sesame/csm-1b" }, "links": { "huggingface": "https://huggingface.co/sesame/csm-1b", "github": "https://github.com/SesameAILabs/csm" }, "hf_downloads_30d": 129861, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "chatterbox-multilingual", "name": "Chatterbox (Multilingual)", "org": "Resemble AI", "type": "TTS", "released": { "value": "2025-04-24", "source": "https://huggingface.co/api/models/ResembleAI/chatterbox", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "500M", "source": "https://github.com/resemble-ai/chatterbox" }, "license": { "value": "mit", "source": "https://huggingface.co/ResembleAI/chatterbox" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/ResembleAI/chatterbox", "note": "MIT; outputs carry Perth neural watermark" }, "languages": { "value": "23+ languages", "source": "https://github.com/resemble-ai/chatterbox" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://github.com/resemble-ai/chatterbox", "note": "Zero-shot, cross-language" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/ResembleAI/chatterbox" }, "links": { "huggingface": "https://huggingface.co/ResembleAI/chatterbox", "github": "https://github.com/resemble-ai/chatterbox" }, "hf_downloads_30d": 1718719, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "chatterbox-turbo", "name": "Chatterbox-Turbo", "org": "Resemble AI", "type": "TTS", "released": { "value": "2025-12-02", "source": "https://huggingface.co/api/models/ResembleAI/chatterbox-turbo", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "350M", "source": "https://huggingface.co/ResembleAI/chatterbox-turbo" }, "license": { "value": "mit", "source": "https://huggingface.co/ResembleAI/chatterbox-turbo" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/ResembleAI/chatterbox-turbo", "note": "MIT; watermarked outputs" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/ResembleAI/chatterbox-turbo" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://huggingface.co/ResembleAI/chatterbox-turbo", "note": "Requires a reference clip" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/ResembleAI/chatterbox-turbo" }, "links": { "huggingface": "https://huggingface.co/ResembleAI/chatterbox-turbo", "github": "https://github.com/resemble-ai/chatterbox" }, "hf_downloads_30d": 0, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "fish-s2-pro", "name": "Fish Audio S2 Pro", "org": "Fish Audio", "type": "TTS", "released": { "value": "2026-03-09", "source": "https://huggingface.co/api/models/fishaudio/s2-pro", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "4B slow AR + 400M fast AR", "source": "https://huggingface.co/fishaudio/s2-pro" }, "license": { "value": "Fish Audio Research License", "source": "https://huggingface.co/fishaudio/s2-pro" }, "commercial_use": { "value": "no", "source": "https://huggingface.co/fishaudio/s2-pro", "note": "Research/non-commercial free; commercial needs a separate license from Fish Audio" }, "languages": { "value": [ "zh", "en", "ja", "ko", "es", "pt", "ar", "ru", "fr", "de", "sv", "it", "tr", "no", "nl", "cy", "eu", "ca", "da", "gl", "ta", "hu", "fi", "pl", "et", "hi", "la", "ur", "th", "vi", "jw", "bn", "yo", "sl", "cs", "sw", "nn", "he", "ms", "uk", "id", "kk", "bg", "lv", "my", "tl", "sk", "ne", "fa", "af", "el", "bo", "hr", "ro", "sn", "mi", "yi", "am", "be", "km", "is", "az", "sd", "br", "sq", "ps", "mn", "ht", "ml", "sr", "sa", "te", "ka", "bs", "pa", "lt", "kn", "si", "hy", "mr", "as", "gu", "fo" ], "source": "https://huggingface.co/fishaudio/s2-pro" }, "streaming": { "value": true, "source": "https://huggingface.co/fishaudio/s2-pro" }, "voice_cloning": { "value": true, "source": "https://github.com/fishaudio/fish-speech", "note": "10-30 s reference audio" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/fishaudio/s2-pro" }, "links": { "huggingface": "https://huggingface.co/fishaudio/s2-pro", "github": "https://github.com/fishaudio/fish-speech" }, "hf_downloads_30d": 35543, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "fish-drama-3", "name": "Fish Audio Drama 3", "org": "Fish Audio", "type": "TTS", "released": { "value": "2026-10-01", "source": "https://huggingface.co/api/models/fishaudio/drama-3", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "35B MoE (3B active per token)", "source": "https://huggingface.co/fishaudio/drama-3" }, "license": { "value": "Fish Audio Community License", "source": "https://huggingface.co/fishaudio/drama-3/blob/main/LICENSE", "note": "HF metadata tag says fish-audio-research-license; card text and LICENSE say Community License" }, "commercial_use": { "value": "conditional", "source": "https://huggingface.co/fishaudio/drama-3/blob/main/LICENSE", "note": "Free commercial use only below an annual revenue threshold; hosted services need a separate license" }, "languages": { "value": [ "zh", "en", "ja", "ko", "es", "pt", "ar", "ru", "fr", "de", "sv", "it", "tr", "no", "nl", "cy", "eu", "ca", "da", "gl", "ta", "hu", "fi", "pl", "et", "hi", "la", "ur", "th", "vi", "jw", "bn", "yo", "sl", "cs", "sw", "nn", "he", "ms", "uk", "id", "kk", "bg", "lv", "my", "tl", "sk", "ne", "fa", "af", "el", "bo", "hr", "ro", "sn", "mi", "yi", "am", "be", "km", "is", "az", "sd", "br", "sq", "ps", "mn", "ht", "ml", "sr", "sa", "te", "ka", "bs", "pa", "lt", "kn", "si", "hy", "mr", "as", "gu", "fo" ], "source": "https://huggingface.co/fishaudio/drama-3" }, "streaming": { "value": true, "source": "https://huggingface.co/fishaudio/drama-3", "note": "<100 ms first audio on one H200 (vendor figure)" }, "voice_cloning": { "value": true, "source": "https://huggingface.co/fishaudio/drama-3" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/fishaudio/drama-3" }, "links": { "huggingface": "https://huggingface.co/fishaudio/drama-3", "github": "https://github.com/fishaudio/fish-speech" }, "hf_downloads_30d": 11795, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "kyutai-tts-1.6b", "name": "Kyutai TTS 1.6B (EN/FR)", "org": "Kyutai", "type": "TTS", "released": { "value": "2025-06-30", "source": "https://huggingface.co/api/models/kyutai/tts-1.6b-en_fr", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "1.8B", "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr", "note": "Card: actually 1.8B, not 1.6B" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr", "note": "CC-BY-4.0" }, "languages": { "value": [ "en", "fr" ], "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr" }, "streaming": { "value": true, "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr" }, "voice_cloning": { "value": "limited", "source": "https://huggingface.co/kyutai/tts-1.6b-en_fr", "note": "Restricted to pre-computed voice embeddings" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/kyutai/tts-1.6b-en_fr" }, "links": { "huggingface": "https://huggingface.co/kyutai/tts-1.6b-en_fr", "github": "https://github.com/kyutai-labs/delayed-streams-modeling" }, "hf_downloads_30d": 85503, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "pocket-tts", "name": "Pocket TTS", "org": "Kyutai", "type": "TTS", "released": { "value": "2025-12-29", "source": "https://huggingface.co/api/models/kyutai/pocket-tts", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "100M", "source": "https://github.com/kyutai-labs/pocket-tts" }, "license": { "value": "cc-by-4.0", "source": "https://huggingface.co/kyutai/pocket-tts" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/kyutai/pocket-tts", "note": "CC-BY-4.0 per HF metadata (repo gated); bundled voices have their own licenses (kyutai/tts-voices)" }, "languages": { "value": [ "en", "fr", "de", "pt", "it", "es", "nl" ], "source": "https://github.com/kyutai-labs/pocket-tts" }, "streaming": { "value": true, "source": "https://github.com/kyutai-labs/pocket-tts" }, "voice_cloning": { "value": true, "source": "https://github.com/kyutai-labs/pocket-tts" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://github.com/kyutai-labs/pocket-tts", "note": "Runs on CPU using 2 cores; ~6x real-time on MacBook Air M4 (vendor figure)" }, "gated_on_hf": { "value": true, "source": "https://huggingface.co/api/models/kyutai/pocket-tts" }, "links": { "huggingface": "https://huggingface.co/kyutai/pocket-tts", "github": "https://github.com/kyutai-labs/pocket-tts" }, "hf_downloads_30d": 1111, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "higgs-tts-3-4b", "name": "Higgs TTS 3 4B", "org": "Boson AI", "type": "TTS", "released": { "value": "2026-06-04", "source": "https://huggingface.co/api/models/bosonai/higgs-tts-3-4b", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "4.65B", "source": "https://huggingface.co/api/models/bosonai/higgs-tts-3-4b", "note": "safetensors parameter count" }, "license": { "value": "Boson Higgs TTS 3 Research and Non-Commercial License", "source": "https://huggingface.co/bosonai/higgs-tts-3-4b" }, "commercial_use": { "value": "conditional", "source": "https://huggingface.co/bosonai/higgs-tts-3-4b", "note": "Non-commercial, plus a Creator Use Grant for monetized content; products/APIs need a commercial license" }, "languages": { "value": "100+ languages", "source": "https://huggingface.co/bosonai/higgs-tts-3-4b" }, "streaming": { "value": true, "source": "https://huggingface.co/bosonai/higgs-tts-3-4b", "note": "SSE streaming via SGLang-Omni / vLLM-Omni" }, "voice_cloning": { "value": true, "source": "https://huggingface.co/bosonai/higgs-tts-3-4b" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/bosonai/higgs-tts-3-4b" }, "links": { "huggingface": "https://huggingface.co/bosonai/higgs-tts-3-4b", "github": "https://github.com/sgl-project/sglang-omni" }, "hf_downloads_30d": 120143, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "vibevoice-realtime-0.5b", "name": "VibeVoice-Realtime 0.5B", "org": "Microsoft", "type": "TTS", "released": { "value": "2025-12-04", "source": "https://huggingface.co/api/models/microsoft/VibeVoice-Realtime-0.5B", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "0.5B LLM (1.02B total in safetensors)", "source": "https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B" }, "license": { "value": "mit", "source": "https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B", "note": "MIT, but Microsoft does not recommend commercial/real-world use without further testing; card lists prohibited uses" }, "languages": { "value": "English (+ experimental voices in 9 more languages)", "source": "https://github.com/microsoft/VibeVoice" }, "streaming": { "value": true, "source": "https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B", "note": "Streaming text input" }, "voice_cloning": { "value": null, "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/microsoft/VibeVoice-Realtime-0.5B" }, "links": { "huggingface": "https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B", "github": "https://github.com/microsoft/VibeVoice" }, "hf_downloads_30d": 293242, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "kitten-tts-mini-0.8", "name": "KittenTTS Mini 0.8", "org": "KittenML", "type": "TTS", "released": { "value": "2026-02-19", "source": "https://huggingface.co/api/models/KittenML/kitten-tts-mini-0.8", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "80M", "source": "https://huggingface.co/KittenML/kitten-tts-mini-0.8", "note": "~79 MB file" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/KittenML/kitten-tts-mini-0.8" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/KittenML/kitten-tts-mini-0.8", "note": "Apache-2.0" }, "languages": { "value": null, "source": null }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": null, "source": null }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://github.com/KittenML/KittenTTS", "note": "Lightweight ONNX, runs on CPU" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/KittenML/kitten-tts-mini-0.8" }, "links": { "huggingface": "https://huggingface.co/KittenML/kitten-tts-mini-0.8", "github": "https://github.com/KittenML/KittenTTS" }, "hf_downloads_30d": 40249, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "qwen3-tts-1.7b-base", "name": "Qwen3-TTS 1.7B Base", "org": "Alibaba Qwen", "type": "TTS", "released": { "value": "2026-01-21", "source": "https://huggingface.co/api/models/Qwen/Qwen3-TTS-12Hz-1.7B-Base", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "1.93B", "source": "https://huggingface.co/api/models/Qwen/Qwen3-TTS-12Hz-1.7B-Base", "note": "safetensors parameter count" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base", "note": "Apache-2.0" }, "languages": { "value": [ "zh", "en", "ja", "ko", "de", "fr", "ru", "pt", "es", "it" ], "source": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base" }, "streaming": { "value": true, "source": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base", "note": "Vendor claims end-to-end latency as low as 97 ms" }, "voice_cloning": { "value": true, "source": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/Qwen/Qwen3-TTS-12Hz-1.7B-Base" }, "links": { "huggingface": "https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base", "github": "https://github.com/QwenLM/Qwen3-TTS" }, "hf_downloads_30d": 3027358, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "supertonic-3", "name": "Supertonic 3", "org": "Supertone", "type": "TTS", "released": { "value": "2026-04-29", "source": "https://github.com/supertone-inc/supertonic", "note": "README news entry" }, "params": { "value": "~99M", "source": "https://huggingface.co/Supertone/supertonic-3" }, "license": { "value": "OpenRAIL-M (weights), MIT (sample code)", "source": "https://huggingface.co/Supertone/supertonic-3" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/Supertone/supertonic-3", "note": "OpenRAIL-M allows commercial use with use-based restrictions" }, "languages": { "value": "31 languages", "source": "https://github.com/supertone-inc/supertonic" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": false, "source": "https://github.com/supertone-inc/supertonic", "note": "Open model uses preset voices; zero-shot cloning is offered in Supertone's paid API" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://huggingface.co/Supertone/supertonic-3", "note": "Does not require a GPU" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/Supertone/supertonic-3" }, "links": { "huggingface": "https://huggingface.co/Supertone/supertonic-3", "github": "https://github.com/supertone-inc/supertonic" }, "hf_downloads_30d": 50227, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "neutts-air", "name": "NeuTTS Air", "org": "Neuphonic", "type": "TTS", "released": { "value": "2025-09-15", "source": "https://huggingface.co/api/models/neuphonic/neutts-air", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "~552M (emb + active), ~360M active", "source": "https://github.com/neuphonic/neutts-air" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/neuphonic/neutts-air" }, "commercial_use": { "value": "yes", "source": "https://github.com/neuphonic/neutts-air", "note": "Air is Apache-2.0; newer Nano variants use NeuTTS Open License 1.0. HF repo gated" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/neuphonic/neutts-air" }, "streaming": { "value": true, "source": "https://github.com/neuphonic/neutts-air", "note": "Streaming via GGUF backbones + llama-cpp-python" }, "voice_cloning": { "value": true, "source": "https://github.com/neuphonic/neutts-air", "note": "From ~3 s of audio" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": true, "source": "https://github.com/neuphonic/neutts-air", "note": "CPU benchmarks published (Ryzen 9HX 370, M4, Galaxy A25)" }, "gated_on_hf": { "value": true, "source": "https://huggingface.co/api/models/neuphonic/neutts-air" }, "links": { "huggingface": "https://huggingface.co/neuphonic/neutts-air", "github": "https://github.com/neuphonic/neutts-air" }, "hf_downloads_30d": 1832, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "indextts-2", "name": "IndexTTS-2", "org": "bilibili (IndexTeam)", "type": "TTS", "released": { "value": "2025-06-18", "source": "https://huggingface.co/api/models/IndexTeam/IndexTTS-2", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": null, "source": null }, "license": { "value": "bilibili Model Use License Agreement", "source": "https://github.com/index-tts/index-tts" }, "commercial_use": { "value": "conditional", "source": "https://github.com/index-tts/index-tts", "note": "Commercial use: contact indexspeech@bilibili.com" }, "languages": { "value": [ "zh", "en" ], "source": "https://huggingface.co/IndexTeam/IndexTTS-2", "note": "IndexTTS-2.5 adds ja, es, ar" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://github.com/index-tts/index-tts", "note": "Single reference clip" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/IndexTeam/IndexTTS-2" }, "links": { "huggingface": "https://huggingface.co/IndexTeam/IndexTTS-2", "github": "https://github.com/index-tts/index-tts" }, "hf_downloads_30d": 13434, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "cosyvoice3-0.5b", "name": "Fun-CosyVoice3 0.5B (2512)", "org": "Alibaba FunAudioLLM", "type": "TTS", "released": { "value": "2025-12-11", "source": "https://huggingface.co/api/models/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "0.5B", "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "note": "from model name" }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "note": "Apache-2.0" }, "languages": { "value": [ "zh", "en", "fr", "es", "ja", "ko", "it", "ru", "de" ], "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512" }, "streaming": { "value": true, "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "note": "Bi-streaming (text in/audio out), ~150 ms latency claimed" }, "voice_cloning": { "value": true, "source": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "note": "Multilingual / cross-lingual zero-shot" }, "min_memory": { "value": null, "gb": null, "kind": null, "source": null }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/FunAudioLLM/Fun-CosyVoice3-0.5B-2512" }, "links": { "huggingface": "https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512", "github": "https://github.com/FunAudioLLM/CosyVoice" }, "hf_downloads_30d": 242469, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "luxtts", "name": "LuxTTS", "org": "YatharthS", "type": "TTS", "released": { "value": "2026-01-22", "source": "https://huggingface.co/api/models/YatharthS/LuxTTS", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": null, "source": null }, "license": { "value": "apache-2.0", "source": "https://huggingface.co/YatharthS/LuxTTS" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/YatharthS/LuxTTS", "note": "Model and code Apache-2.0" }, "languages": { "value": [ "en" ], "source": "https://huggingface.co/YatharthS/LuxTTS" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://huggingface.co/YatharthS/LuxTTS", "note": "Min 3 s reference" }, "min_memory": { "value": "Fits within 1 GB VRAM", "gb": 1, "kind": "VRAM", "source": "https://huggingface.co/YatharthS/LuxTTS" }, "runs_on_cpu": { "value": true, "source": "https://github.com/ysharma3501/LuxTTS", "note": "Faster than real-time on CPU (vendor claim)" }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/YatharthS/LuxTTS" }, "links": { "huggingface": "https://huggingface.co/YatharthS/LuxTTS", "github": "https://github.com/ysharma3501/LuxTTS" }, "hf_downloads_30d": 5942, "benchmarks": [], "benchmark_status": "benchmark pending" }, { "id": "miso-tts-8b", "name": "Miso TTS 8B", "org": "Miso Labs", "type": "TTS", "released": { "value": "2026-05-21", "source": "https://huggingface.co/api/models/MisoLabs/MisoTTS", "note": "Hugging Face repo creation date (proxy for release date)" }, "params": { "value": "~8.2B", "source": "https://github.com/MisoLabsAI/MisoTTS" }, "license": { "value": "Modified MIT", "source": "https://huggingface.co/MisoLabs/MisoTTS/blob/main/LICENSE" }, "commercial_use": { "value": "yes", "source": "https://huggingface.co/MisoLabs/MisoTTS/blob/main/LICENSE", "note": "Products with >50M MAU or >$10M monthly revenue must display 'Miso Labs'" }, "languages": { "value": [ "en" ], "source": "https://github.com/MisoLabsAI/MisoTTS" }, "streaming": { "value": null, "source": null }, "voice_cloning": { "value": true, "source": "https://github.com/MisoLabsAI/MisoTTS", "note": "Conditions on prior audio context" }, "min_memory": { "value": "24 GB VRAM recommended (bf16); ~20 GB RAM on CPU (slow)", "gb": 24, "kind": "VRAM", "source": "https://github.com/MisoLabsAI/MisoTTS" }, "runs_on_cpu": { "value": null, "source": null }, "gated_on_hf": { "value": false, "source": "https://huggingface.co/api/models/MisoLabs/MisoTTS" }, "links": { "huggingface": "https://huggingface.co/MisoLabs/MisoTTS", "github": "https://github.com/MisoLabsAI/MisoTTS" }, "hf_downloads_30d": 0, "benchmarks": [], "benchmark_status": "benchmark pending" } ] }