This topic describes the billing details for Tongyi Tingwu.
Free trial
New users who activate the service can enjoy a 90-day free trial. No fees are charged while the service status is Free trial version.
Service capability | Free trial benefits |
Real-time recording |
|
Audio and video file recording |
|
Billing for new APIs
Commercial pricing for new APIs
Module | Feature | Optional | Standard price (CNY/hour) |
ASR | Real-time meeting transcription (includes speaker diarization) | Choose one | 0.6 |
Audio and video file transcription (includes speaker diarization and automatic language detection) | 0.6 | ||
Large Language Model (LLM) | Includes: Chapter overviews, full-text summaries, speaker summaries, Q&A reviews, mind maps, action items, keywords, key points, spoken-to-written text conversion, and custom prompts | Each feature is optional and billed independently based on the duration of the processed audio or video. If you use multiple features or multiple prompts, the fees are cumulative. For example, if you use keywords and chapter overviews, the fee is 0.064 + 0.064 = 0.128 CNY/hour. | 0.064 |
Includes: Service quality inspection and conversation content extraction | Optional. Billed based on the duration of the processed audio or video. | 0.13 | |
Multimodal | Multimodal (PPT extraction from video + PPT summary) | Optional | 0.64 |
Translation | Real-time translation | Optional | 4 |
Offline translation | Optional | 0.5 |
Billing rules
Transcription: Tongyi Tingwu real-time meeting recording supports up to three audio ingest endpoints. If you input only one audio stream, you are charged for the total audio duration. For example, if you push an audio stream of silence or pure noise for one hour, you are still charged for one hour of transcription. When you use two or three ingest endpoints, you are charged only for the duration that produces transcription results. However, if no ingest endpoint produces transcription results at any given time, you are still charged. Because multiple people do not often speak at the same time in a meeting, the billed duration is typically equal to or slightly longer than the meeting duration. Audio duration is accumulated in seconds. For example, if you upload three audio files of 1,000 s, 3,000 s, and 500 s, the total audio duration is 1,000 + 3,000 + 500 = 4,500 s. The Automated Speech Recognition (ASR) fee is calculated as (4,500 / 3,600) × 0.6 = 0.75 CNY.
Large Language Model (LLM): Fees for LLM features such as chapter overviews, full-text summaries, and keywords are cumulative. Each feature you use is billed separately. If you submit multiple custom prompts in a single request, you are charged for each prompt.
Translation: If you enable translation for a real-time meeting, you are charged for both real-time recording and real-time translation. For example, the fee for a one-hour real-time recording with translation for one ingest endpoint is 0.6 + 4 = 4.6 CNY. If you use bilingual (Chinese and English) transcription and set both Chinese and English as target languages for translation, the translation fee is doubled. For example, the fee for a one-hour bilingual transcription with one ingest endpoint translated into both Chinese and English is 0.6 + (4 × 2) = 8.6 CNY.
Commercial pricing for legacy APIs (higher rates, switch to the new version as soon as possible)
Tongyi Tingwu supports only the pay-as-you-go billing method for legacy APIs. Fees are settled daily, with tiered pricing based on usage.
Product Specifications | Standard price | Daily usage | Tiered unit price (CNY/hour) |
Real-time recording | 10.5 CNY/hour | 0 to 299 hours | 10.5 |
300 to 999 hours | 9 | ||
1,000 to 2,999 hours | 7.2 | ||
3,000 to 4,999 hours | 4.8 | ||
5,000 hours or more | 3.6 | ||
Audio and video file recording | 9.5 CNY/hour | 0 to 299 hours | 9.5 |
300 to 999 hours | 8.2 | ||
1,000 to 2,999 hours | 6.6 | ||
3,000 to 4,999 hours | 4.7 | ||
5,000 hours or more | 3.4 | ||
Real-time translation | 8 CNY/hour | 0 to 299 hours | 8 |
300 to 999 hours | 7.5 | ||
1,000 to 2,999 hours | 7 | ||
3,000 to 4,999 hours | 6.5 | ||
5,000 hours or more | 6 | ||
Offline translation | 1.2 CNY/hour | 0 to 299 hours | 0.9 |
300 to 999 hours | 0.8 | ||
1,000 to 2,999 hours | 0.7 | ||
3,000 to 4,999 hours | 0.6 | ||
5,000 hours or more | 0.5 |