使用unsloth 的 8-bit GGUF 量化版，llama.cpp 部署， Qwen 3.6 27B 和 35B A3B MoE 部署本地电脑，相当于GPT-5 或 Claude Sonnet 4.5

        llama.cpp 部署环境而非 Ollama，服务端一行命令：llama-server -hf unsloth/Qwen3.6-27B-MTP-GGUF:Q8_0 --spec-type draft-mtp -ngl 999 -fa on -c 65536，OpenCode 挂上去直接写代码

llama-server -hf unsloth/Qwen3.6-27B-MTP-GGUF:Q8_0 \
    --spec-type draft-mtp -ngl 999 -fa on -c 65536 --port 8080

What it does:

• -hf unsloth/Qwen3.6-27B-MTP-GGUF:Q8_0 grabs from Hugging Face, on the next runs will reuse that
• -m ~/models/Qwen3.6-27B-Q8_0.gguf use instead if you already have it
• draft-mtp we use a fast model to predict subsequent tokens, speeds up things
• -ngl 999 for putting all layers to GPU
• -fa on flash attention is on
• -c 65536 context size set to 64k tokens (this we can tweak, as Qwen 3.6 27B native context is 256k)
• --port 8080 better to pin port, as it will be used by other configs

If you open http://127.0.0.1:8080, you can directly chat with it.

配置~/.config/opencode/opencode.jsonc 文件

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "llama": {
      "name": "llama.cpp (local)",
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "http://127.0.0.1:8080/v1",
        "apiKey": "local"
      },
      "models": {
        "qwen3.6-27b": { "name": "Qwen3.6-27B Q8 +MTP" }
      }
    }
  },
  "model": "llama/qwen3.6-27b"
}

llama-cli -hf unsloth/Qwen3.6-27B-MTP-GGUF:Q8_0 \
    -ngl 999 -fa on -c 65536

参考

Qwen 3.6 27B is the sweet spot for local development - Quesma Blog