Dear Users, To continuously enhance our large model service capabilities, we will update the list of supported models on the platform in alignment with official schedules from major model providers. Please review the following adjustments and prepare your applications accordingly.
Models to Be Deprecated
The following models will be officially discontinued on March 31, 2026. Please migrate to alternative models before this date:
- GPT: gpt-4o, gpt-4o-mini
- Gemini: gemini-2.0-flash
- Doubao: doubao-seed-1.6
Newly Added Models
We are simultaneously introducing the latest high-performance models from multiple providers to offer you more options:
- GPT: gpt-5.2, gpt-5.1, gpt-5, gpt-5-mini, gpt-5-nano
- Gemini: gemini-3.0-pro, gemini-3.0-flash
- Doubao: doubao-seed-1.8
- Qwen: qwen3-max, qwen-plus, qwen-flash
Migration Guidance
There are no exact one-to-one replacements for the deprecated models. The recommendations below are based on similar performance or use-case alignment. We strongly advise thorough functional and quality testing before switching models to ensure service stability and user experience:
- For gpt-4o → recommended replacement: gpt-5-mini
- For gpt-4o-mini → recommended replacement: gpt-5-mini
- For gemini-2.0-flash → recommended replacement: gemini-3.0-flash (note: slightly higher latency; if this does not meet your requirements, consider gpt-5-mini)
- For doubao-seed-1.6 → recommended replacement: doubao-seed-1.8
To prevent disruption to your live services, any calls to deprecated models after March 31, 2026, that have not been manually updated will be automatically redirected to the recommended replacement models listed above.
| If you do not wish to receive such emails, please click here to unsubscribe. |
