export LITELLM_API_KEY="sk-1234"
npm install -g @openai/codex@nativemodel = "gemini-2.5-pro"
model_provider = "litellm"
[model_providers.litellm]
name = "LiteLLM"
base_url = "https://litellm.example.com"
env_key = "LITELLM_API_KEY"
wire_api = "responses"$HOME/.codex/config.toml
export GOOGLE_GEMINI_BASE_URL="https://litellm.example.com"
export GEMINI_API_KEY="sk-1234"
export GEMINI_MODEL=gemini-2.5-pro
npm install -g @ai-codespark/gemini-cli@latest{
"selectedAuthType": "gemini-api-key",
"theme": "ANSI"
}$HOME/.gemini/settings.json
export SILICONFLOW_BASE_URL="https://litellm.example.com"
export SILICONFLOW_API_KEY="sk-1234"
export GEMINI_MODEL=gemini-2.5-pro
npm install -g @gen-cli/gen-cli@latest{
"selectedAuthType": "siliconflow-api-key"
}$HOME/.gen-cli/settings.json
export KIMI_VERSION=1.14.0
export KIMI_BASE_URL="https://litellm.example.com"
export KIMI_API_KEY=sk-1234
export KIMI_MODEL_NAME=gemini-2.5-pro
curl -L https://github.com/MoonshotAI/kimi-cli/releases/download/${KIMI_VERSION}/kimi-${KIMI_VERSION}-x86_64-unknown-linux-gnu.tar.gz -o /tmp/kimi.tar.gz && \
tar -xzf /tmp/kimi.tar.gz -C /tmp && \
mv /tmp/kimi /usr/local/bin/kimi && \
chown root:root /usr/local/bin/kimi && \
chmod +x /usr/local/bin/kimi && \
rm -rf /tmp/kimi.tar.gzdefault_model = "kimi-for-coding"
default_thinking = false
default_yolo = false
[providers.kimi-for-coding]
type = "kimi"
base_url = "https://api.kimi.com/coding/v1"
api_key = "sk-xxx"
[models.kimi-for-coding]
provider = "kimi-for-coding"
model = "kimi-for-coding"
max_context_size = 262144$HOME/.kimi/config.toml
BUN_INSTALL="$HOME/.bun"
PATH="$BUN_INSTALL/bin:$PATH"
XDG_DATA_HOME="$HOME/.local/share"
XDG_CACHE_HOME="$HOME/.cache"
XDG_CONFIG_HOME="$HOME/.config"
XDG_STATE_HOME="$HOME/.local/state"
mkdir -p \
$HOME/.local/share/opencode/bin \
$HOME/.local/share/opencode/log \
$HOME/.cache/opencode \
$HOME/.config/opencode \
$HOME/.local/state/opencode
bun install -g @ai-sdk/openai-compatible && \
bun install -g opencode-ai
mkdir -p /tmp/test-opencode && \
cd /tmp/test-opencode && \
git init && \
(echo "hello" | opencode "$PWD" > /tmp/opencode.log 2>&1 &) && \
OPENCODE_PID=$! && \
sleep 5 && \
kill $OPENCODE_PID 2>/dev/null || true && \
rm -rf /tmp/test-opencode
echo "export TERM=xterm-256color" >> /home/$NAME/.bashrc{
"$schema": "https://opencode.ai/config.json",
"model": "litellm/ollama-kimi-k2.5",
"provider": {
"litellm": {
"npm": "@ai-sdk/openai-compatible",
"name": "LiteLLM",
"options": {
"baseURL": "https://litellm.example.com",
"apiKey": "sk-1234"
},
"models": {
"ollama-kimi-k2.5": {
"name": "Ollama Kimi K2.5"
}
}
}
}
}$HOME/.config/opencode/opencode.jsonc
export OPENAI_BASE_URL="https://litellm.example.com"
export OPENAI_API_KEY="sk-1234"
export OPENAI_MODEL="gemini-2.5-pro"
npm install -g @qwen-code/qwen-code@latest{
"selectedAuthType": "openai"
}$HOME/.qwen/settings.json
uv venv
uv sync --all-extras
source .venv/bin/activate{
"default_provider": "openai",
"max_steps": 20,
"enable_lakeview": false,
"model_providers": {
"openai": {
"api_key": "sk-1234",
"base_url": "https://litellm.example.com",
"model": "gemini-2.5-pro",
"max_tokens": 120000,
"temperature": 0.5,
"top_p": 1,
"max_retries": 10
}
}
}trae_config.json
