Speech-to-text

Canary-Qwen 2.5B requirements, license & speed

By NVIDIA · released 2025-06-26 · 2.56B parameters

Hugging Face ↗ GitHub ↗ Compare

Commercial useYes
Min memoryNot stated
Runs on CPUUnknown
StreamingUnknown

Key facts

License
cc-by-4.0 source
Commercial use
Yes source
CC-BY-4.0; card: ready for commercial use
Parameters
2.56B source
safetensors parameter count
Minimum memory
Not stated by source
Runs on CPU
Unknown
Streaming
Unknown
Voice cloning
n/a (speech-to-text)
Languages
en source
Released
2025-06-26 source
Hugging Face repo creation date (proxy for release date)
Gated on HF
No

Measured speed & memory

Benchmark pending. We haven't measured Canary-Qwen 2.5B yet. Planned targets: Raspberry Pi 5, CPU-only x86, Apple Silicon Mac and consumer GPUs. Get notified.

Accuracy

We don't run accuracy evals. For word error rate, check the Hugging Face Open ASR Leaderboard.

Similar STT models

Spotted an error? Report it. Every fact links to the model card or repo it came from.

New open speech models: will they run on your hardware?

One email a week: new open TTS/STT models, license gotchas, and what we measured on CPUs, Macs, Pis and consumer GPUs.