ComfyUI Node
🦙 Llama.cpp Router Client (Chat + Vision + Audio + Video + Stream)
A ComfyUI node in LLM / Simple llama.cpp with 17 inputs and 5 outputs.
🦙 Llama.cpp Router Client (Chat + Vision + Audio + Video + Stream)
- image
- response
- thinking
- json
- raw
- model_used
â—„server_urlhttp://127.0.0.1:8888â–º
â—„modelâ–º
â—„system_promptYou are a helpful assistant.â–º
â—„promptHello!â–º
â—„streamtrueâ–º
â—„max_tokens0â–º
â—„temperature0.80â–º
â—„seed-1â–º
â—„json_modefalseâ–º
â—„unload_after_generatefalseâ–º
â—„text_postprocessfix_mojibakeâ–º
â—„audio_pathâ–º
â—„video_pathâ–º
â—„api_keyâ–º
â—„json_schema_hintâ–º
â—„stop_customâ–º
CategoryLLM / Simple llama.cpp
Inputs (17)
| Name | Type | Default | Description |
|---|---|---|---|
| server_url | STRING | http://127.0.0.1:8888 | llama.cpp server base URL (router mode or single-model) |
| model | STRING | Model name — click Fetch Models button to pick from the server | |
| system_prompt | STRING | You are a helpful assistant. | System prompt sent before the user message |
| prompt | STRING | Hello! | User message / question |
| stream | BOOLEAN | true | Enable SSE streaming for real-time token output |
| max_tokens | INT | 00–131072 | Max tokens to generate (0 = server default) |
| temperature | FLOAT | 0.800–2 | Sampling temperature |
| seed | INT | -1-1–2147483647 | RNG seed (-1 = random) |
| json_mode | BOOLEAN | false | Force JSON-only output via response_format |
| unload_after_generate | BOOLEAN | false | Call /models/unload on the server after generation |
| text_postprocess | COMBO | fix_mojibake | Text post-processing mode |
| imageopt | IMAGE | Optional image for vision models (first frame used) | |
| audio_pathopt | STRING | Optional path to audio file (WAV/MP3) for audio models | |
| video_pathopt | STRING | Optional path to video file for video models | |
| api_keyopt | STRING | API key for Bearer token auth (leave empty if none) | |
| json_schema_hintopt | STRING | JSON schema hint appended to system prompt when json_mode is on | |
| stop_customopt | STRING | Custom stop string (leave empty for none) |
Outputs (5)
| Name | Type | Description |
|---|---|---|
| response | STRING | Clean response text with thinking blocks removed |
| thinking | STRING | Extracted thinking/reasoning content (empty if model produced none) |
| json | STRING | Parsed JSON object (empty if json_mode is off or no JSON found) |
| raw | STRING | Full raw server response as JSON string |
| model_used | STRING | The model name that was actually used for this request |