Best free NVIDIA Canary-Qwen 2.5B alternatives

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Open-source speech-to-text, run locally free.

• Open-source speech-to-text, run locally free
• 99+ languages and translation
• Robust to accents and noise
• Powers countless apps

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Tops the Hugging Face Open ASR Leaderboard with 6.05% average WER at 0.6B params.

• Tops the Hugging Face Open ASR Leaderboard with 6.05% average WER at 0.6B params
• CC-BY-4.0 license, free for commercial and non-commercial use
• RTFx ~3,386 — extremely fast inference relative to accuracy

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — 70x real-time transcription speed via batched inference on top of Whisper.

• 70x real-time transcription speed via batched inference on top of Whisper
• Adds wav2vec2 forced alignment for accurate word-level timestamps Whisper lacks natively
• Built-in speaker diarization and VAD preprocessing to cut hallucinations

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Open-source speech recognition toolkit with Paraformer and SenseVoice models.

• Open-source speech recognition toolkit with Paraformer and SenseVoice models
• Streaming ASR, VAD, punctuation and speaker diarization
• Strong Mandarin and multilingual accuracy
• Runs on CPU or GPU, MIT-style license

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — C/C++ port of OpenAI Whisper running on CPU, Metal, Core ML, CUDA and Vulkan.

• C/C++ port of OpenAI Whisper running on CPU, Metal, Core ML, CUDA and Vulkan
• Runs on phones, Raspberry Pi and browsers via WASM
• MIT licensed

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Whisper reimplementation on CTranslate2, up to 4x faster with less memory.

• Whisper reimplementation on CTranslate2, up to 4x faster with less memory
• int8 quantization for GPU and CPU
• MIT licensed Python library

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Small streaming speech-recognition models for phones, browsers and low-power devices.

• Small streaming speech-recognition models for phones, browsers and low-power devices
• Low latency for live voice interfaces
• Free open-source weights and runtime

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Offline speech-to-text, text-to-speech, VAD and speaker diarization without internet.

• Offline speech-to-text, text-to-speech, VAD and speaker diarization without internet
• Runs on Android, iOS, Raspberry Pi, WebAssembly and desktop with many language bindings
• Apache-2.0

✦ WhyA NVIDIA Canary-Qwen 2.5B alternative — Fast, accurate speech-to-text API.

• Fast, accurate speech-to-text API
• Real-time streaming and diarization
• Voice-agent building blocks
• Free credits to start