Providers
Supercode supports multiple AI providers. You can configure them through the CLI with /connect or through the web dashboard at the Providers page.
Currently supported providers:
- Google Gemini
- OpenRouter
- MiniMax
- NVIDIA NIM
- OpenAI
- Anthropic
Provider Configuration
CLI: /connect
The /connect command walks you through configuring a provider interactively:
/connect
Select from the list of supported providers, then paste your API key when prompted. Keys are sent to the server and stored securely.
Web Dashboard
The Providers page in the dashboard offers a visual interface:
- Navigate to the Providers page from the sidebar
- Each provider card shows its status (connected/disconnected)
- Click a provider to enter or update your API key
- Set a default provider for new sessions
Google Gemini
Google Gemini offers competitive pricing and strong performance.
| Model | Description | |-------|-------------| | gemini-2.5-flash | Fast, cost-effective for most tasks | | gemini-2.0-flash | Previous generation, widely available |
Setup:
- Get an API key from Google AI Studio
- Configure via
/connector Dashboard - Server must have
GOOGLE_GENERATIVE_AI_API_KEYset
OpenRouter
OpenRouter provides access to multiple models through a single API.
| Model | Description | |-------|-------------| | gpt-oss-120b | Open-source 120B parameter model | | deepseek-v4-flash | Fast reasoning model | | claude-3.5-sonnet | Anthropic's balanced model | | Various | Access to 100+ models |
Setup:
- Create an account at openrouter.ai
- Generate an API key
- Server must have
OPENROUTER_API_KEYset
MiniMax
MiniMax provides the MiniMax-M2 model with strong coding capabilities.
| Model | Description | |-------|-------------| | MiniMax-M2 | General purpose coding model |
Setup:
- Sign up at minimax.com
- Generate an API key
- Server must have
MINIMAX_API_KEYset
NVIDIA NIM
NVIDIA NIM offers optimized model inference with multiple model options.
| Model | Description | |-------|-------------| | minimax-m2.7 | MiniMax optimized for NVIDIA | | deepseek-v4-flash | DeepSeek reasoning model | | llama-3.3-70b | Meta's 70B parameter model |
Setup:
- Access NVIDIA NIM at build.nvidia.com
- Generate an API key
- Server must have
NVIDIA_API_KEYset
OpenAI
OpenAI provides GPT-4 and GPT-4o models.
| Model | Description | |-------|-------------| | gpt-4o | Latest multimodal model | | gpt-4o-mini | Smaller, faster variant | | gpt-4-turbo | High-performance option |
Setup:
- Get an API key from platform.openai.com
- Configure via
/connector Dashboard
Anthropic
Anthropic provides Claude models with strong reasoning capabilities.
| Model | Description | |-------|-------------| | claude-4-opus | Most capable model | | claude-4-sonnet | Balanced performance | | claude-3.5-haiku | Fast, lightweight |
Setup:
- Get an API key from anthropic.com
- Configure via
/connector Dashboard
Environment Variables
For self-hosted or development setups, configure providers via environment variables:
# Google Gemini
GOOGLE_GENERATIVE_AI_API_KEY=your-key
# OpenRouter
OPENROUTER_API_KEY=your-key
# MiniMax
MINIMAX_API_KEY=your-key
MINIMAX_API_BASE=https://api.minimax.com
# NVIDIA NIM
NVIDIA_API_KEY=your-key
NVIDIA_API_BASE=https://api.nvcf.nvidia.com
# OpenAI
OPENAI_API_KEY=your-key
# Anthropic
ANTHROPIC_API_KEY=your-key
Switching Providers Mid-Session
Use the /model command to switch providers or models during an active session:
/model
This opens a selection menu. The conversation history is preserved when switching.
Default Provider
Set a default provider so new sessions always start with your preferred model:
- CLI: Set via the initial setup prompts
- Dashboard: Select default in the Providers page
Troubleshooting
API Key Issues
Error: Provider API key not configured
- Run
/connectto set up a provider - Verify the API key in the Dashboard
- Ensure the API key has not expired
- Check server environment variables for correct configuration
Rate Limiting
If you encounter rate limits:
- Wait before retrying
- Consider switching to a different provider
- Upgrade your API plan
Model Not Available
Model unavailable for this provider
- Check that the provider supports the requested model
- Verify your API plan includes access to the model
- Try a different model from the same provider