| Model | Per minute | Strength |
|---|---|---|
| hy-asr-3.0-preview (Tencent Hunyuan) | $0.003 | Accurate transcription, timestamps + SRT, strong on Chinese |
| wand-asr-v1 (Tencent WAND) | $0.007 | Multilingual + automatic language detection |
| NovAI (Hunyuan ASR) | OpenAI Whisper | |
|---|---|---|
| Price per minute | $0.003 | $0.006 |
| Price per hour | $0.18 | $0.36 |
| Timestamps | Yes (word / sentence) | Yes (verbose_json) |
| SRT subtitle output | Built-in URL | Assemble yourself |
| Chinese / CJK accuracy | Native-strong | Good |
Source: OpenAI lists whisper-1 transcription at $0.006/min ($0.36/hour). NovAI’s ASR is token-billed at $1.54/1M upstream, reconciled to real audio duration; the per-minute figures above are the live references.
curl https://aiapi-pro.com/v1/audio/transcriptions \
-H "Authorization: Bearer $NOVAI_API_KEY" \
-F model=hy-asr-3.0-preview \
-F [email protected]
# -> { "text": "...", "segments": [ { "start": 0.0, "end": 4.2, "text": "..." } ],
# "srt_url": "https://.../subtitles.srt" }The endpoint mirrors OpenAI’s /v1/audio/transcriptions, so existing Whisper client code often works by changing only the base URL and the model name.
NovAI gives you OpenAI-compatible access to 85 Chinese frontier models — video, image, 3D, audio and chat — on a prepaid balance with no subscription and zero platform fee. $2 free credit on signup, no card required.
Start Free →