# Speech-to-Text and Transcription API

> Compare transcription and ASR models by exact endpoint mode, audio-hour pricing, inputs, transcript output, and limitations.

- Canonical: https://chinaapi.ai/speech-to-text-api/
- Updated: 2026-08-08
- Tasks: transcription
- Endpoint: `POST /v1/audio/transcriptions or POST /v1/chat/completions with input_audio, according to api_mode`
- Inputs: An audio file or base64 input_audio payload, depending on the model's api_mode.
- Outputs: Transcript text or a structured transcription response.
- Lifecycle: Synchronous for the listed public endpoint modes. Persist the transcript and handle rejected file formats explicitly.
- Limits: File size, duration, diarization, timestamps, dialect coverage, and accepted formats are unknown unless explicitly listed for the model.

## Models with matching catalog metadata

### [stepaudio-2-asr-pro](https://chinaapi.ai/models/stepaudio-2-asr-pro/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/audio/transcriptions`
- Cost example: 1 audio hour × $0.35 = $0.35.

### [step-asr](https://chinaapi.ai/models/step-asr/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/audio/transcriptions`
- Cost example: 1 audio hour × $0.15 = $0.15.

### [step-asr-1.1](https://chinaapi.ai/models/step-asr-1.1/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/audio/transcriptions`
- Cost example: 1 audio hour × $0.35 = $0.35.

### [mimo-v2.5-asr](https://chinaapi.ai/models/mimo-v2.5-asr/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/chat/completions`
- Cost example: 1 audio hour × $0.074 = $0.074.

### [qwen3-asr-flash](https://chinaapi.ai/models/qwen3-asr-flash/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/audio/transcriptions`
- Cost example: 1 audio hour × $0.123539 = $0.123539.

### [stepaudio-2.5-asr](https://chinaapi.ai/models/stepaudio-2.5-asr/)

- Tasks: video-with-audio, transcription
- Endpoint: `POST https://api.chinaapi.ai/v1/audio/transcriptions`
- Cost example: 1 audio hour × $0.022 = $0.022.

Displayed prices are generated from `scripts/models-data.json`, synchronized against the ChinaAPI gateway. Media service margin, if any, is included in the displayed USD rate and is not added separately.

Source: https://chinaapi.ai/data/model-pricing/ · updated 2026-08-08.