Skip to main content

Providers

Suzent is model-agnostic. Configure any number of providers in Settings → Providers and switch between them per session.

API keys are stored in the local database — never in plain text config files.


Configuring a Provider

  1. Open Settings → Providers
  2. Select the provider you want to add
  3. Enter your credentials and click Save
  4. Click Verify to confirm the connection and load available models
  5. Enable the models you want to use from the model list

Cloud Providers

OpenAI

Models: GPT-4.1, GPT-4.1 Mini, o3, o4-mini

Get your key: platform.openai.com/api-keys

FieldDescription
OPENAI_API_KEYYour API key (starts with sk-...)
OPENAI_BASE_URLOptional. Override the endpoint — useful for Azure OpenAI or compatible proxies

Anthropic

Models: Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku 4.5

Get your key: console.anthropic.com/settings/keys

FieldDescription
ANTHROPIC_API_KEYYour API key (starts with sk-ant-...)

Google Gemini

Models: Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.0 Flash

Get your key (free tier available): aistudio.google.com/app/apikey

FieldDescription
GEMINI_API_KEYYour API key (starts with AIza...)

GOOGLE_API_KEY is also accepted as an alias.


xAI (Grok)

Models: Grok 3, Grok 3 Mini, Grok 3 Fast

Get your key: console.x.ai

FieldDescription
XAI_API_KEYYour API key (starts with xai-...)

DeepSeek

Models: DeepSeek V3 (Chat), DeepSeek R1 (Reasoner)

Get your key: platform.deepseek.com/api_keys

FieldDescription
DEEPSEEK_API_KEYYour API key (starts with sk-...)

DeepSeek R1 is a reasoning model — slower but stronger for multi-step problems.


MiniMax

Models: MiniMax M2.5, MiniMax M2.1

Get your key: platform.minimaxi.com

FieldDescription
MINIMAX_API_KEYYour API key

Moonshot (Kimi)

Models: Kimi v1 128K, Kimi v1 32K, Kimi v1 8K

Get your key: platform.moonshot.cn

FieldDescription
MOONSHOT_API_KEYYour API key (starts with sk-...)

Zhipu AI (GLM)

Models: GLM-4.7 Flash

Get your key: open.bigmodel.cn

FieldDescription
ZAI_API_KEYYour API key

Aggregators & Proxies

OpenRouter

Access 300+ models from a single API key — GPT, Claude, Gemini, Mistral, Llama, and more. Useful if you want to experiment across providers without managing separate subscriptions.

Get your key: openrouter.ai/keys

FieldDescription
OPENROUTER_API_KEYYour API key (starts with sk-or-...)

After saving, click Fetch Models to load the full catalog.


LiteLLM Proxy

Run a self-hosted gateway in front of any set of models. Useful for teams, rate-limit management, or cost tracking.

Setup: docs.litellm.ai

FieldDescription
LITELLM_MASTER_KEYYour proxy master key (starts with sk-...)
LITELLM_BASE_URLURL of your running proxy (e.g. http://localhost:4000)

Click Fetch Models after saving to load the models your proxy exposes.


Local

Ollama

Run models entirely on your machine — no API key or internet connection needed.

Setup: Install Ollama from ollama.com, then pull a model:

ollama pull llama3.2
FieldDescription
OLLAMA_BASE_URLBase URL of your Ollama server (default: http://localhost:11434)
OLLAMA_API_KEYOptional. Only needed if your instance requires authentication

Click Fetch Models to auto-detect all locally available models.


Tips

Environment variables — You can set provider keys as env vars before launching Suzent instead of using the UI. The app checks the database first, then falls back to the environment. Keys set via env show up in Settings as "Set in env" and cannot be overwritten from the UI.

Custom models — For each provider you can add custom model IDs on top of the default list. In Settings → Providers → [Provider] → Custom Models, enter the model ID in LiteLLM format: provider/model-name (e.g. openai/gpt-4o-2024-11-20).