Batch Processing
The Batch API allows you to process large volumes of requests asynchronously with significant cost savings.
Overview
Batch processing is ideal for scenarios where you need to process many requests but don't require immediate responses. You can save up to 50% on both input and output tokens compared to real-time API calls.
Pricing
Batch processing offers a 50% discount on all token usage:
| Model | Batch Input | Batch Output |
|---|---|---|
| RakutenAI-3.0 | $0.25/M | $1.00/M |
| RakutenAI-2.0 (8x7B) | $0.025/M | $0.10/M |
| RakutenAI-7B | $0.015/M | $0.06/M |
| RakutenAI-2.0-mini | $0.01/M | $0.04/M |
For complete pricing information, see our pricing guide.
How it works
Detailed documentation on how to use the Batch API will be provided here, including:
- Creating batch jobs
- Uploading request files
- Monitoring batch status
- Retrieving results
- Error handling
Use cases
Common use cases for batch processing will be documented here, such as:
- Large-scale content generation
- Data analysis and classification
- Bulk translation tasks
- Evaluation and testing
Get Started with Batch Processing
Ready to start processing large volumes of requests with cost savings?
Try in Playground