[Bug]: Using DeepSeek with LlamaIndex causes a 400 Bad Request because LlamaIndex calls the deprecated /v1/completions text endpoint instead
![[Bug]: Using DeepSeek with LlamaIndex causes a 400 Bad Request because LlamaIndex calls the deprecated /v1/completions text endpoint instead](https://www.chat-gpts.plus/wp-content/uploads/2026/08/22846-c859ce1c-768x403.jpg)
当 LlamaIndex 配合 DeepSeek 模型使用 OpenAILike 时,会因默认调用已废弃的 /v1/completions 文本补全端点而返回 400 Bad Request。优先排查是否显式设置了 is_chat_model=True ,并确认 model 与 api_base 参



![[Bug]: vllm 0.23.0 and 0.24.0 - Qwen3.6-35B-A3B-FP8 - Fails generating code- "400 Unterminated string starting at"](https://www.chat-gpts.plus/wp-content/uploads/2026/08/47761-b6f2eca2-768x403.jpg)




![[Bug]: Current main Docker compose fails to start with v0.25.6 image](https://www.chat-gpts.plus/wp-content/uploads/2026/08/15692-ba3ab69e-768x403.jpg)