LLM Providers

Suitcase's setup dashboard (/setup) now follows Hamr's model connection flow:

1. OAuth login — OpenAI/Codex and Claude/Anthropic where Hamr exposes OAuth.
2. Provider API key — select a provider and paste only its API key; no endpoint is required.
3. Custom compatible endpoint — use the old OpenAI/Anthropic-compatible base URL + model form for local servers, relays, and private gateways.

Provider API keys are saved through Hamr auth. Custom endpoint credentials are saved in Suitcase's local secret store.

Provider API keys

Use Use a provider API key for OpenAI, Claude/Anthropic, Google Gemini, OpenRouter, DeepSeek, Groq, Mistral, xAI, Together, or another Hamr catalog provider. Pick provider, paste key, choose model, save.

No base URL is needed in this mode.

Custom endpoint modes

OpenAI (default)

When LLM_PROVIDER=openai, Suitcase auto-configures task-scoped models:

Task	Default Model	Purpose
Chat	gpt-4o	Normal conversation
Fast	gpt-4o-mini	Triage, lightweight extraction
Trusted	gpt-4o	Application generation, reviews, bootstrap
Vision	gpt-4o	PDF visual review

Base URL defaults to https://api.openai.com/v1.

OpenRouter

Pre-configured with free models from OpenRouter:

Task	Default Model
Chat / Trusted / Vision	moonshotai/kimi-k2.6 (free)
Fast (triage)	google/gemma-4-31b-it:free (free)

Base URL defaults to https://openrouter.ai/api/v1. Kimi K2.6 is currently free — just bring your sk-or-... API key.

OpenAI-compatible

Use this for any OpenAI-compatible endpoint, including LM Studio, Ollama-compatible proxies, DeepSeek, Groq, Together, or a private relay. Local endpoints can leave LLM_API_KEY empty.

LLM_PROVIDER=openai-compatible
LLM_BASE_URL=https://your-endpoint/v1
LLM_MODEL=your-model-name
LLM_API_KEY=optional-for-local

Anthropic-compatible

One model for all tasks. Use for Anthropic's API or a compatible proxy.

LLM_PROVIDER=anthropic-compatible
LLM_BASE_URL=https://api.anthropic.com
LLM_MODEL=claude-sonnet-4-20250514

Quick Reference

Provider	Base URL	Mode
OpenAI	https://api.openai.com/v1	openai
OpenRouter	https://openrouter.ai/api/v1	openrouter
Anthropic	https://api.anthropic.com	anthropic-compatible
LM Studio	http://127.0.0.1:1234/v1	openai-compatible
Ollama	http://127.0.0.1:11434/v1	openai-compatible
DeepSeek	https://api.deepseek.com/v1	openai-compatible
Groq	https://api.groq.com/openai/v1	openai-compatible
Together	https://api.together.xyz/v1	openai-compatible
Custom	https://your-server/v1	openai-compatible

Configuration

Configure from the Setup dashboard:

1. Click Login with OpenAI OAuth or Login with Claude OAuth, if you want OAuth.
2. Or click Use a provider API key, select provider, paste API key, choose model, and save.
3. Or click Custom OpenAI/Anthropic-compatible endpoint for base URL + model configuration.
4. Click save; Suitcase verifies the selected Hamr/provider path.

Vision Support

For resume PDF visual review, use a model with vision support:

• OpenAI mode: gpt-4o is used by default (vision-capable)
• OpenRouter mode: moonshotai/kimi-k2.6 supports vision
• Compatible modes: configure LLM_VISION_MODEL in .env