ai-workspace-infra/playbooks
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(openclaw): adopt native provider simulation pointing to litellm gateway

Browse Source
This commit is contained in:
Haitao Pan 2026-06-23 12:42:04 +08:00
parent 25b8204b7b
commit a0d59c0af1
2 changed files with 99 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

108
roles/vhosts/gateway_openclaw/defaults/main.yml
View File

@ -56,27 +56,43 @@ gateway_openclaw_acp_max_concurrent_sessions: 2
gateway_openclaw_acp_backend: acpx
gateway_openclaw_acp_default_agent: codex
gateway_openclaw_codex_app_server_url: ws://127.0.0.1:9001
gateway_openclaw_default_model_primary: "deepseek-v4-flash"
gateway_openclaw_default_model_fallback: "deepseek-v4-pro"
gateway_openclaw_default_model_primary: "deepseek/deepseek-v4-flash"
gateway_openclaw_default_model_fallback: "deepseek/deepseek-v4-pro"
gateway_openclaw_default_model:
primary: "{{ gateway_openclaw_default_model_primary }}"
fallbacks:
- "{{ gateway_openclaw_default_model_fallback }}"
- "deepseek-chat"
- "deepseek-reasoner"
- "nvidia-glm-5.2"
- "nvidia-minimax-m3"
- "nvidia-qwen3.5"
- "ollama-cloud-kimi-k2.7-code"
- "deepseek/deepseek-chat"
- "deepseek/deepseek-reasoner"
- "nvidia/deepseek-v4-flash"
- "nvidia/deepseek-v4-pro"
- "nvidia/glm-5.2"
- "nvidia/minimax-m3"
- "nvidia/qwen3.5"
- "nvidia/kimi-k2.7-code"
- "ollama/deepseek-v4-flash"
- "ollama/deepseek-v4-pro"
- "ollama/glm-5.2"
- "ollama/minimax-m3"
- "ollama/qwen3.5"
- "ollama/kimi-k2.7-code"
gateway_openclaw_default_models:
"{{ gateway_openclaw_default_model_primary }}": {}
"{{ gateway_openclaw_default_model_fallback }}": {}
"deepseek-chat": {}
"deepseek-reasoner": {}
"nvidia-glm-5.2": {}
"nvidia-minimax-m3": {}
"nvidia-qwen3.5": {}
"ollama-cloud-kimi-k2.7-code": {}
"deepseek/deepseek-chat": {}
"deepseek/deepseek-reasoner": {}
"nvidia/deepseek-v4-flash": {}
"nvidia/deepseek-v4-pro": {}
"nvidia/glm-5.2": {}
"nvidia/minimax-m3": {}
"nvidia/qwen3.5": {}
"nvidia/kimi-k2.7-code": {}
"ollama/deepseek-v4-flash": {}
"ollama/deepseek-v4-pro": {}
"ollama/glm-5.2": {}
"ollama/minimax-m3": {}
"ollama/qwen3.5": {}
"ollama/kimi-k2.7-code": {}
openai/gpt-5.5:
agentRuntime:
id: codex
@ -128,7 +144,7 @@ gateway_openclaw_mcp_servers:
transport: streamable-http
gateway_openclaw_model_providers:
litellm:
deepseek:
api: openai-completions
baseUrl: "http://127.0.0.1:{{ litellm_listen_port | default(4000) }}/v1"
apiKey: "{{ ai_workspace_auth_token }}"
@ -145,28 +161,70 @@ gateway_openclaw_model_providers:
contextWindow: 128000
maxTokens: 8192
reasoning: true
- id: "deepseek-chat"
- id: "deepseek/deepseek-chat"
name: DeepSeek V3 Chat
input: [text]
contextWindow: 64000
- id: "deepseek-reasoner"
- id: "deepseek/deepseek-reasoner"
name: DeepSeek R1 Reasoner
input: [text]
contextWindow: 64000
reasoning: true
- id: "nvidia-glm-5.2"
name: GLM 5.2
nvidia:
api: openai-completions
baseUrl: "http://127.0.0.1:{{ litellm_listen_port | default(4000) }}/v1"
apiKey: "{{ ai_workspace_auth_token }}"
models:
- id: "nvidia/deepseek-v4-flash"
name: NVIDIA DeepSeek V4 Flash
input: [text]
contextWindow: 128000
- id: "nvidia-minimax-m3"
name: MiniMax M3
- id: "nvidia/deepseek-v4-pro"
name: NVIDIA DeepSeek V4 Pro
input: [text]
contextWindow: 128000
- id: "nvidia-qwen3.5"
name: Qwen 3.5
- id: "nvidia/glm-5.2"
name: NVIDIA GLM 5.2
input: [text]
contextWindow: 128000
- id: "ollama-cloud-kimi-k2.7-code"
name: Kimi K2.7 Code
- id: "nvidia/minimax-m3"
name: NVIDIA MiniMax M3
input: [text]
contextWindow: 128000
- id: "nvidia/qwen3.5"
name: NVIDIA Qwen 3.5
input: [text]
contextWindow: 128000
- id: "nvidia/kimi-k2.7-code"
name: NVIDIA Kimi K2.7 Code
input: [text]
contextWindow: 128000
ollama:
api: openai-completions
baseUrl: "http://127.0.0.1:{{ litellm_listen_port | default(4000) }}/v1"
apiKey: "{{ ai_workspace_auth_token }}"
models:
- id: "ollama/deepseek-v4-flash"
name: Ollama DeepSeek V4 Flash
input: [text]
contextWindow: 128000
- id: "ollama/deepseek-v4-pro"
name: Ollama DeepSeek V4 Pro
input: [text]
contextWindow: 128000
- id: "ollama/glm-5.2"
name: Ollama GLM 5.2
input: [text]
contextWindow: 128000
- id: "ollama/minimax-m3"
name: Ollama MiniMax M3
input: [text]
contextWindow: 128000
- id: "ollama/qwen3.5"
name: Ollama Qwen 3.5
input: [text]
contextWindow: 128000
- id: "ollama/kimi-k2.7-code"
name: Ollama Kimi K2.7 Code
input: [text]
contextWindow: 128000

24
roles/vhosts/litellm/files/register_mainstream_models.sh
View File

@ -89,19 +89,22 @@ if [ -n "${DEEPSEEK_API_KEY:-}" ]; then
echo "========================================="
echo "Registering DeepSeek Models..."
echo "========================================="
add_model "deepseek-v4-flash" "deepseek/deepseek-v4-flash" "DEEPSEEK_API_KEY"
add_model "deepseek-v4-pro" "deepseek/deepseek-v4-pro" "DEEPSEEK_API_KEY"
add_model "deepseek-chat" "deepseek/deepseek-chat" "DEEPSEEK_API_KEY"
add_model "deepseek-reasoner" "deepseek/deepseek-reasoner" "DEEPSEEK_API_KEY"
add_model "deepseek/deepseek-v4-flash" "deepseek/deepseek-v4-flash" "DEEPSEEK_API_KEY"
add_model "deepseek/deepseek-v4-pro" "deepseek/deepseek-v4-pro" "DEEPSEEK_API_KEY"
add_model "deepseek/deepseek-chat" "deepseek/deepseek-chat" "DEEPSEEK_API_KEY"
add_model "deepseek/deepseek-reasoner" "deepseek/deepseek-reasoner" "DEEPSEEK_API_KEY"
fi
if [ -n "${NVIDIA_API_KEY:-}" ]; then
echo "========================================="
echo "Registering NVIDIA Build Models..."
echo "========================================="
add_model "nvidia-glm-5.2" "openai/thudm/glm-5.2-chat" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia-minimax-m3" "openai/minimax/minimax-m3" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia-qwen3.5" "openai/alibaba/qwen3.5-72b-instruct" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/deepseek-v4-flash" "openai/deepseek-v4-flash" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/deepseek-v4-pro" "openai/deepseek-v4-pro" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/glm-5.2" "openai/thudm/glm-5.2-chat" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/minimax-m3" "openai/minimax/minimax-m3" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/qwen3.5" "openai/alibaba/qwen3.5-72b-instruct" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
add_model "nvidia/kimi-k2.7-code" "openai/moonshot/kimi-k2.7-code" "NVIDIA_API_KEY" "https://integrate.api.nvidia.com/v1"
fi
echo "========================================="
@ -136,7 +139,12 @@ if [ -n "${OLLAMA_API_KEY:-}" ]; then
echo "Registering OLLAMA Cloud Models..."
echo "========================================="
OLLAMA_API_BASE="${OLLAMA_API_BASE:-https://api.ollama.cloud/v1}"
add_model "ollama-cloud-kimi-k2.7-code" "openai/moonshot/kimi-k2.7-code" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/deepseek-v4-flash" "openai/deepseek-v4-flash" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/deepseek-v4-pro" "openai/deepseek-v4-pro" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/glm-5.2" "openai/thudm/glm-5.2-chat" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/minimax-m3" "openai/minimax/minimax-m3" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/qwen3.5" "openai/alibaba/qwen3.5-72b-instruct" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
add_model "ollama/kimi-k2.7-code" "openai/moonshot/kimi-k2.7-code" "OLLAMA_API_KEY" "$OLLAMA_API_BASE"
fi
echo "All models requested have been registered."