Hugging Face

Hugging Face provides inference access to popular open-source models through their Inference Providers ecosystem. Many models are available for free, making it an excellent option for experimentation and budget-conscious development.

Website: https://huggingface.co/

Getting an API Key

1. Sign Up/Sign In: Go to Hugging Face. Create an account or sign in.
2. Navigate to Settings: Go to your Access Tokens page.
3. Create a Token: Generate a new access token with appropriate permissions.
4. Copy the Token: Copy the token immediately and store it securely.

Supported Models

Cline supports the following Hugging Face models (all currently free):

• moonshotai/Kimi-K2-Instruct (Default) - Advanced reasoning model with 131K context, superior coding and math capabilities
• openai/gpt-oss-120b - Large 120B open-weight reasoning model for complex tasks (131K context)
• openai/gpt-oss-20b - Medium 20B open-weight model balancing reasoning with accessibility (131K context)
• deepseek-ai/DeepSeek-V3-0324 - Advanced reasoning model (64K context)
• deepseek-ai/DeepSeek-R1 - DeepSeek's reasoning model with step-by-step thinking (64K context)
• deepseek-ai/DeepSeek-R1-0528 - Latest DeepSeek reasoning version (64K context)
• meta-llama/Llama-3.1-8B-Instruct - Efficient 8B Llama model for general tasks (128K context)

Configuration in Cline

1. Open Cline Settings: Click the settings icon (⚙️) in the Cline panel.
2. Select Provider: Choose "Hugging Face" from the "API Provider" dropdown.
3. Enter Token: Paste your Hugging Face access token.
4. Select Model: Choose your desired model from the "Model" dropdown.

Tips and Notes

• Free Inference: All listed models are currently available at no cost through Hugging Face Inference Providers.
• Open Source: All models are open-source and can also be self-hosted.
• Rate Limits: Free tier has rate limits. Check Hugging Face documentation for current limits.
• Model Availability: Model availability depends on inference provider capacity and may vary.