> ## Documentation Index
> Fetch the complete documentation index at: https://docs.novita.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Sep 29 and Oct 9, 2026 Model Deprecation Notice

> Novita will retire 34 models in two batches on Sep 29 and Oct 9, 2026, covering the DeepSeek, Qwen, GLM, MiniMax, and ERNIE series. Migrate to the recommended replacement model before each cutoff.

The following models will be retired from the Novita platform in two batches, on Sep 29 and Oct 9, 2026. If your application uses an affected model, migrate to the recommended replacement model before its retirement date. A `—` in the replacement column means the model has no direct replacement.

## Affected models

| Model ID | Retirement date | Replacement model |
| - | - | - |
| `deepseek/deepseek-v3.2` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `deepseek/deepseek-r1` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `deepseek/deepseek-v3.1` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `deepseek/deepseek-v3.1-terminus` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `deepseek/deepseek-v3.2-exp` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `deepseek/deepseek-r1-distill-llama-70b` | Sep 29, 2026 | — |
| `deepseek/deepseek-v3/community` | Sep 29, 2026 | `deepseek/deepseek-v4.1-flash` |
| `qwen/qwen3.6-plus` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-vl-235b-a22b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `openai/gpt-oss-20b` | Oct 9, 2026 | — |
| `deepseek/deepseek-v4-flash-vision-exp` | Oct 9, 2026 | `zai-org/glm-5.3-flash` |
| `qwen/qwen3-coder-30b-a3b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-coder-480b-a35b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-coder-next` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen-2.5-72b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-next-80b-a3b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3.5-plus` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `sao10k/l31-70b-euryale-v2.2` | Oct 9, 2026 | — |
| `qwen/qwen3-vl-235b-a22b-thinking` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-max` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `mistralai/mistral-nemo` | Oct 9, 2026 | — |
| `sao10k/l3-8b-lunaris` | Oct 9, 2026 | — |
| `zai-org/glm-5v-turbo` | Oct 9, 2026 | `zai-org/glm-5.3-flash` |
| `qwen/qwen3-235b-a22b-thinking-2507` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-235b-a22b-instruct-2507` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `baidu/ernie-4.5-vl-424b-a47b` | Oct 9, 2026 | — |
| `minimax/minimax-m2` | Oct 9, 2026 | `minimax/minimax-m3` |
| `qwen/qwen3-omni-30b-a3b-instruct` | Oct 9, 2026 | `qwen/qwen3.8-flash` |
| `minimax/minimax-m2.1` | Oct 9, 2026 | `minimax/minimax-m3` |
| `zai-org/glm-5-turbo` | Oct 9, 2026 | `zai-org/glm-5.3-flash` |
| `minimax/minimax-m2.5-highspeed` | Oct 9, 2026 | `minimax/minimax-m3` |
| `qwen/qwen3-235b-a22b-fp8` | Oct 9, 2026 | `qwen/qwen3.8-max` |
| `qwen/qwen3-embedding-0.6b` | Oct 9, 2026 | — |
| `thudm/glm-4-32b-0414` | Oct 9, 2026 | `zai-org/glm-5.3-flash` |

## Action required

If your application uses an affected model, migrate to the recommended replacement model before its retirement date. For models with no direct replacement, remove the affected calls from your application before that date.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.