A comprehensive configuration skill to integrate SiliconFlow's inference platform into OpenClaw for access to over 98 AI models including DeepSeek, Qwen, and Kimi.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install add-siliconflow-provider
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install add-siliconflow-provider using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is SiliconFlow Provider Configuration?
This configuration skill allows developers to integrate SiliconFlow as a primary model source within the OpenClaw environment. SiliconFlow is a premier AI model inference platform that provides a standard OpenAI-compatible protocol, making it easy to swap or supplement existing providers. By utilizing this addition to your Openclaw Skills, you gain access to a massive library of over 98 chat models, including flagship series like DeepSeek-V3, Qwen3, and GLM-4.
The primary value of this skill lies in its ability to significantly lower inference costs—offering flagship models at 30-50% of official prices—while also providing access to permanently free models like Qwen3-8B and DeepSeek-R1-8B. This makes it an essential component for developers building cost-efficient and resilient AI workflows through Openclaw Skills.
SiliconFlow Provider Configuration Use Cases
- Accessing free-tier models for local development and testing without incurring costs.
- Implementing a fallback strategy where the system switches to SiliconFlow when other providers are unavailable.
- Consolidating multiple Chinese LLM providers (DeepSeek, Kimi, GLM, MiniMax) under a single API key.
- Deploying cost-sensitive AI agents that leverage SiliconFlow's low-priced flagship model inference.
How SiliconFlow Provider Configuration Works
- Retrieve a secure API key from the SiliconFlow dashboard.
- Create a backup of the existing OpenClaw configuration file to ensure system stability.
- Patch the openclaw.json file to register the SiliconFlow provider with its base URL and API protocol.
- Define the model list within the configuration, including specific parameters like context windows and cost structures.
- Create human-readable aliases for complex model IDs to simplify chat commands.
- Integrate the new models into the global fallback chain for improved reliability.
- Restart the OpenClaw gateway and verify the connection using the built-in diagnostic tools.
SiliconFlow Provider Configuration Setup
First, obtain your API key from the SiliconFlow console. Then, back up your current configuration:
cp ~/.openclaw/openclaw.json ~/.openclaw/openclaw.json.backup.$(date +%Y%m%d_%H%M%S)
Apply the SiliconFlow provider patch to your openclaw.json using the gateway config.patch command or manual editing. Ensure the baseUrl is set to https://api.siliconflow.cn/v1 and the api protocol is set to openai-completions. After updating the config, restart your services:
openclaw gateway restart
openclaw doctor
SiliconFlow Provider Configuration Data Schema & Taxonomy
The skill organizes provider data and model metadata in a structured format within the main configuration file. This allows Openclaw Skills to manage cost tracking and model selection effectively.
| Object Key | Type | Description |
|---|---|---|
providers.siliconflow |
Object | Contains API key, base URL, and protocol definitions. |
models |
Array | A list of supported model IDs, context windows, and token costs. |
agents.defaults.models |
Object | Mapping of full provider IDs to short aliases (e.g., sf-qwen3-8b). |
fallbacks |
Array | Priority list for model failover including SiliconFlow endpoints. |
SiliconFlow Provider Configuration Advanced Features
- Support for reasoning models: Integrated support for DeepSeek-R1 and other reasoning-capable LLMs.
- Pro Tier Access: Configuration options for 'Pro/' prefixed models using dedicated high-speed inference clusters.
- Dynamic Cost Mapping: Real-time cost tracking per million tokens for both input and output.
- Multi-Agent Aliasing: Custom model aliases that allow agents to switch between 480B MoE models and lightweight 8B models instantly.
SKILL.md
Loading
Related Openclaw Skills
NewCLI Model Provider Configuration
A configuration skill for integrating NewCLI as a multi-model provider for OpenClaw, featuring high-efficiency AWS routes for Claude models.
MiniMax Provider Configuration
A technical skill for integrating MiniMax M2.1 series models into OpenClaw using both direct API Key and OAuth portal access methods.
Spacesuit Framework for OpenClaw
A batteries-included framework scaffold designed to standardize and secure AI agent workspaces.
OpenClaw Command Center
A mission control dashboard providing real-time session monitoring and cost intelligence for your AI agent workforce.
Callmac: Remote Voice Control for Mac
Control your Mac's voice remotely from mobile devices to play announcements, alarms, and stories using high-quality neural TTS.
OpenClaw Starter Guide
A strategic roadmap for deploying, configuring, and optimizing cost-effective multi-agent AI systems from zero to professional scale.
Featured*
