ASR Channels
Speech Recognition (ASR) is the first step of video translation — converting speech from audio or video into timed subtitle files. pyVideoTrans supports 15+ recognition channels.
Local Offline Recognition
No internet connection required after initial model download. Runs completely offline.
| Provider | Description | GPU Accelerated | Rating |
|---|---|---|---|
| faster-whisper (Local Built-in) | Fast, high quality, supports dozens of languages | ✅ | ⭐⭐⭐ Default recommendation |
| openai-whisper (Local Built-in) | High accuracy, slower | ✅ | ⭐⭐⭐ |
| Qwen-ASR (Local Built-in) | Excellent for Chinese | ✅ | ⭐⭐⭐ Recommended for Chinese |
| FunASR-Chinese (Local Built-in) | By Alibaba, optimized for Chinese | ✅ | ⭐⭐⭐ Recommended for Chinese |
| Huggingface_ASR (Local Built-in) | Supports multiple language models | ✅ | ⭐⭐ |
| Faster-Whisper-XXL.exe | Windows extra-large model version | ✅ | ⭐⭐ |
| whisper.cpp | Uses whisper.cpp backend | ✅ | ⭐⭐ |
| Parakeet-tdt (Local API) | NVIDIA open-source model | ✅ | ⭐⭐ |
| STT (Local API) | Requires self-deployment | — | ⭐ |
| Whisper.NET | Supports AMD GPUs | ✅ | ⭐⭐ |
Model Selection Guide
| Model | Speed | Accuracy | VRAM Required |
|---|---|---|---|
| tiny | Fastest | Low | ~1GB |
| base | Fast | Low-Medium | ~1GB |
| small | Medium | Medium | ~2GB |
| medium | Slow | High | ~5GB |
| large-v3 | Slowest | Highest | ~8GB |
| large-v3-turbo | Fairly fast | High | ~6GB |
Recommended: large-v3-turbo — best balance of speed and quality.
Online Recognition (Free Tier Available)
| Provider | Description |
|---|---|
| Alibaba Bailian Qwen3-ASR | Requires Alibaba Bailian platform service |
| Elevenlabs.io | Limited free quota |
| Deepgram.com | High accuracy; requires API key registration |
| Gemini AI | Strong for minor languages; requires proxy for users in China |
Online Recognition (Paid)
| Provider | Description |
|---|---|
| 302.AI | Apply at 302.ai |
| ByteDance ASR Speed Edition | Excellent for Chinese |
| OpenAI Speech Recognition API | Excellent quality; requires API key |
Advanced Custom
| Provider | Description |
|---|---|
| Custom ASR API | Build your own recognition interface |
Available Models for Huggingface_ASR
| Model | Supported Languages |
|---|---|
| nvidia/parakeet-ctc-1.1b | English |
| reazon-research/japanese-wav2vec2-large-rs35kh | Japanese |
| kotoba-tech/kotoba-whisper-v2.0 | Japanese |
| biodatlab/whisper-th-large-v3 | Thai |
| vinai/Phowhisper-large | Vietnamese |
| openai/whisper-large-v3 | All languages |