Earlier Kimi frontier model for long-context agents, coding, and multimodal work
https://api-inference.modelscope.cn/v1 moonshotai/Kimi-K2.5 | Benchmark | Score | Metric | Date |
|---|---|---|---|
| SWE-Bench Verified | 70.8 | resolved | Not listed |
| SWE-Atlas Codebase QnA | 13.1 | score | Not listed |
| SWE-Atlas Refactoring | 20.95 | score | Not listed |
| SWE-Atlas Test Writing | 25.77 | score | Not listed |
Kimi K2.5 is listed for chat workloads and supports a 262K context window.
Check ModelScope's current endpoint status before integrating this listing. Use the alternatives table if the provider listing is unavailable.
We only found this Kimi K2.5 listing on ModelScope in the current catalog. Use the related models section for nearby alternatives.
| Provider | Model listing | Access | Context | API | Limits |
|---|---|---|---|---|---|
| | Kimi K2.5 | Check provider | 262K | Native | Varies |
| Model | Provider | Context | Access |
|---|---|---|---|
| Qwen/Qwen3.5-35B-A3B | ModelScope | 131K | Check provider |
| Qwen/Qwen3.5-27B | ModelScope | 131K | Check provider |
| MiniMax-M2.5-highspeed | ModelScope | 205K | Check provider |
| Qwen/Qwen-Image | ModelScope | 131K | Check provider |
| deepseek-ai/DeepSeek-V3.2 | ModelScope | 8K | Check provider |
Kimi K2.5 is tagged for chat in this catalog and works with OpenAI-compatible client libraries.
Kimi K2.5 appears in the free model catalog for ModelScope, but its current endpoint availability should be confirmed with the provider before use.
The model ID shown in this catalog is moonshotai/Kimi-K2.5.
The listed context window is 262K tokens with up to 262K output tokens.
Free Kimi K2.5 API.
For API keys, setup steps, and provider-level limits, see the ModelScope provider page.