- Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathcmdcode-proxy.env
More file actions
Latest commit
19 lines (19 loc) · 684 Bytes
/
Copy pathcmdcode-proxy.env
File metadata and controls
19 lines (19 loc) · 684 Bytes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# cmdcode proxy configuration for vLLM
# Use the proxy as an OpenAI-compatible backend:
#
# Option 1: OpenAI client pointing to proxy
# base_url = "http://127.0.0.1:18080/v1"
# api_key = "not-needed"
#
# Option 2: LiteLLM config (see LiteLLM setup)
# Option 3: Direct API calls
# POST http://127.0.0.1:18080/v1/chat/completions
# Content-Type: application/json
# Authorization: Bearer not-needed
# {"model":"xiaomi/mimo-v2.5","messages":[{"role":"user","content":"Hello"}]}
#
# Supported models (via proxy):
# - xiaomi/mimo-v2.5 (MiMo V2.5)
# - gpt-5.6-luna (GPT-5.6 Luna)
# - claude-sonnet-5 (Claude Sonnet 5)
# - deepseek/deepseek-v4-pro (DeepSeek V4 Pro)