Imported GGUF (hf.co/…) thinks by default, but reports no `thinking` capability and rejects `think: true`

用户在 Ollama 中通过 `ollama pull hf.co/...` 或 `ollama run hf.co/...` 导入并使用 Gemma-4 系列 GGUF 模型。当用户通过 API 或客户端尝试启用 Thinking 功能(设置 `think: true`)时,遇到 HTTP 400






![[FEATURE] Add Wasm-based code interpreter tool to safely run untrusted code.](https://www.chat-gpts.plus/wp-content/uploads/2026/06/4810-3cca9c44-768x403.jpg)

![[Bug]: test_flashinfer_cutlass_mxfp4_fused_moe accuracy mismatch on H20 (sm90) — 89% mismatch vs 20% threshold](https://www.chat-gpts.plus/wp-content/uploads/2026/06/46585-55e7134d-768x403.jpg)