Skip to content

providers.md

Latest commit

 

History

History
1290 lines (1063 loc) · 61.2 KB

providers.md

File metadata and controls

1290 lines (1063 loc) · 61.2 KB
sidebar_position 2
title Configure LLM Provider

import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem'; import { PanelLeft } from 'lucide-react'; import { ModelSelectionTip } from '@site/src/components/ModelSelectionTip'; import { OnboardingProviderSetup } from '@site/src/components/OnboardingProviderSetup';

Supported LLM Providers

goose is compatible with a wide range of LLM providers, allowing you to choose and integrate your preferred model.

:::tip Model Selection Berkeley Function-Calling Leaderboard can be a good guide for selecting models. :::

Available Providers

Provider Description Parameters
Amazon Bedrock Offers a variety of foundation models, including Claude, Jurassic-2, and others. AWS environment variables must be set in advance, not configured through goose configure Credential auth: AWS_PROFILE, or AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_REGION

Bearer token auth: AWS_BEARER_TOKEN_BEDROCK and AWS_REGION, AWS_DEFAULT_REGION, or AWS_PROFILE
Amazon SageMaker TGI Run Text Generation Inference models through Amazon SageMaker endpoints. AWS credentials must be configured in advance. SAGEMAKER_ENDPOINT_NAME, AWS_REGION (optional), AWS_PROFILE (optional)
Anthropic Offers Claude, an advanced AI model for natural language tasks. ANTHROPIC_API_KEY, ANTHROPIC_HOST (optional)
Avian Cost-effective inference API with DeepSeek, Kimi, GLM, and MiniMax models. OpenAI-compatible with streaming and function calling support. AVIAN_API_KEY, AVIAN_HOST (optional)
Azure OpenAI Access Azure-hosted OpenAI models, including GPT-4 and GPT-3.5. Supports both API key and Azure credential chain authentication. AZURE_OPENAI_ENDPOINT, AZURE_OPENAI_DEPLOYMENT_NAME, AZURE_OPENAI_API_KEY (optional)
ChatGPT Codex Access GPT-5 Codex models optimized for code generation and understanding. Requires a ChatGPT Plus/Pro subscription. No manual key. Uses browser-based OAuth authentication for both CLI and Desktop.
Databricks Unified data analytics and AI platform for building and deploying models. DATABRICKS_HOST, DATABRICKS_TOKEN
Docker Model Runner Local models running in Docker Desktop or Docker CE with OpenAI-compatible API endpoints. Because this provider runs locally, you must first download a model. OPENAI_HOST, OPENAI_BASE_PATH
Gemini Advanced LLMs by Google with multimodal capabilities (text, images). Gemini 3 models support configurable thinking levels. GOOGLE_API_KEY, GEMINI3_THINKING_LEVEL (optional)
GCP Vertex AI Google Cloud's Vertex AI platform, supporting Gemini and Claude models. Credentials must be configured in advance. Filters for allowed models by organization policy (if configured). GCP_PROJECT_ID, GCP_LOCATION and optionally GCP_MAX_RATE_LIMIT_RETRIES (5), GCP_MAX_OVERLOADED_RETRIES (5), GCP_INITIAL_RETRY_INTERVAL_MS (5000), GCP_BACKOFF_MULTIPLIER (2.0), GCP_MAX_RETRY_INTERVAL_MS (320_000).
GitHub Copilot Access to AI models from OpenAI, Anthropic, Google, and other providers through GitHub's Copilot infrastructure. GitHub account with Copilot access required. No manual key. Uses device flow authentication for both CLI and Desktop.
Groq High-performance inference hardware and tools for LLMs. GROQ_API_KEY
LiteLLM LiteLLM proxy supporting multiple models with automatic prompt caching and unified API access. LITELLM_HOST, LITELLM_BASE_PATH (optional), LITELLM_API_KEY (optional), LITELLM_CUSTOM_HEADERS (optional), LITELLM_TIMEOUT (optional)
LM Studio Run local models with LM Studio's OpenAI-compatible server. Because this provider runs locally, you must first download a model. None required. Connects to local server at localhost:1234 by default.
Mistral AI Provides access to Mistral models including general-purpose models, specialized coding models (Codestral), and multimodal models (Pixtral). MISTRAL_API_KEY
Ollama Local model runner supporting Qwen, Llama, DeepSeek, and other open-source models. Because this provider runs locally, you must first download and run a model. OLLAMA_HOST
OpenAI Provides gpt-4o, o1, and other advanced language models. Also supports OpenAI-compatible endpoints (e.g., self-hosted LLaMA, vLLM, KServe). o1-mini and o1-preview are not supported because goose uses tool calling. OPENAI_API_KEY, OPENAI_HOST (optional), OPENAI_ORGANIZATION (optional), OPENAI_PROJECT (optional), OPENAI_CUSTOM_HEADERS (optional)
OpenRouter API gateway for unified access to various models with features like rate-limiting management. OPENROUTER_API_KEY
OVHcloud AI Provides access to open-source models including Qwen, Llama, Mistral, and DeepSeek through AI Endpoints service. OVHCLOUD_API_KEY
Ramalama Local model using native OCI container runtimes, CNCF tools, and supporting models as OCI artifacts. Ramalama API is a compatible alternative to Ollama and can be used with the goose Ollama provider. Supports Qwen, Llama, DeepSeek, and other open-source models. Because this provider runs locally, you must first download and run a model. OLLAMA_HOST
Snowflake Access the latest models using Snowflake Cortex services, including Claude models. Requires a Snowflake account and programmatic access token (PAT). SNOWFLAKE_HOST, SNOWFLAKE_TOKEN
VMware Tanzu Platform Enterprise-managed LLM access through AI Services on VMware Tanzu Platform. Models are fetched dynamically from the endpoint. TANZU_AI_API_KEY, TANZU_AI_ENDPOINT
Tetrate Agent Router Service Unified API gateway for AI models including Claude, Gemini, GPT, open-weight models, and others. Supports PKCE authentication flow for secure API key generation. TETRATE_API_KEY, TETRATE_HOST (optional)
Venice AI Provides access to open source models like Llama, Mistral, and Qwen while prioritizing user privacy. Requires an account and an API key. VENICE_API_KEY, VENICE_HOST (optional), VENICE_BASE_PATH (optional), VENICE_MODELS_PATH (optional)
Cerebras Fast inference on Cerebras wafer-scale engines with models like Llama, Qwen, and others. CEREBRAS_API_KEY
xAI Access to xAI's Grok models including grok-3, grok-3-mini, and grok-3-fast with 131,072 token context window. XAI_API_KEY, XAI_HOST (optional)

:::tip Prompt Caching for Claude Models goose automatically enables Anthropic's prompt caching when using Claude models via Anthropic, Amazon Bedrock, Databricks, OpenRouter, and LiteLLM providers. This adds cache_control markers to requests, which can reduce costs for longer conversations by caching frequently-used context. See the provider implementations for technical details. :::

CLI Providers

Provider Description Requirements
Cursor Agent (cursor-agent) Uses Cursor's AI CLI tool with your Cursor subscription. Provides access to GPT-5, Claude 4, and other models through the cursor-agent command-line interface. cursor-agent CLI installed and authenticated

ACP Providers

goose supports Agent Client Protocol (ACP) agents as providers. ACP providers pass goose extensions through to the agent as MCP servers.

Provider Description Requirements
Claude ACP (claude-acp) Uses Claude Code via ACP. Passes goose extensions to the agent as MCP servers. npm install -g @zed-industries/claude-agent-acp, active Claude Code subscription
Codex ACP (codex-acp) Uses OpenAI Codex via ACP. Passes goose extensions to the agent as MCP servers. npm install -g @zed-industries/codex-acp, active ChatGPT Plus/Pro subscription
Gemini ACP (gemini-acp) Uses Google's Gemini CLI via ACP (native --acp support). Passes goose extensions to the agent as MCP servers. npm install -g @google/gemini-cli, authenticated with Google account

:::tip ACP Providers See the ACP Providers guide for detailed setup instructions. :::

Configure Provider and Model

To configure your chosen provider, see available options, or select a model, visit the Models tab in goose Desktop or run goose configure in the CLI.

**First-time users:**

On the welcome screen the first time you open goose, you have these options:

1. Choose `Quick Setup with API Key`. 2. Enter your API key from your provider (for example, OpenAI, Anthropic, or Google). 3. goose will automatically detect your provider and configure the connection. 4. When setup is complete, you're ready to begin your first session. 
<TabItem value="chatgpt" label="ChatGPT Subscription">
1. Choose `ChatGPT Subscription`.
2. goose will open a browser window for you to sign in with the credentials of your active ChatGPT Plus or Pro subscription.
3. Authorize goose to access your ChatGPT subscription.
4. When you return to goose Desktop, you're ready to begin your first session.
</TabItem>
<TabItem value="tetrate" label="Agent Router">
We recommend new users start with Agent Router by Tetrate. Tetrate provides access to multiple AI models with built-in rate limiting and automatic failover. 

:::info Free Credits Offer
You'll receive $10 in free credits the first time you automatically authenticate with Tetrate through goose. This offer is available to both new and existing Tetrate users.
:::
1. Choose `Agent Router by Tetrate`. 
2. goose will open a browser window for you to authenticate with Tetrate, or create a new account if you don't have one already.
3. When you return to goose Desktop, you're ready to begin your first session.
</TabItem>

<TabItem value="openrouter" label="OpenRouter">
1. Choose `Automatic setup with OpenRouter`. 
2. goose will open a browser window for you to authenticate with OpenRouter, or create a new account if you don't have one already.
3. When you return to the goose Desktop, you're ready to begin your first session.
</TabItem>

<TabItem value="others" label="Other Providers">
1. If you have a specific provider you want to use with goose, and an API key from that provider, choose `Other Providers`. 
2. Find the provider of your choice and click its `Configure` button. If you don't see your provider in the list, click `Add Custom Provider` at the bottom of the window to [configure a custom provider](#configure-custom-provider). 
3. Depending on your provider, you'll need to input your API Key, API Host, or other optional [parameters](#available-providers). Click the `Submit` button to authenticate and begin your first session.

:::info Ollama Model Detection
For Ollama users, all locally installed models display automatically in the model selection dropdown.
:::

</TabItem>
**To update your LLM provider and API key:** 1. Click the button in the top-left to open the sidebar 2. Click the `Settings` button on the sidebar 3. Click the `Models` tab 4. Click `Configure providers` 5. Click your provider in the list 6. Add your API key and other required configurations, then click `Submit`

To change your current model:

  1. Click the button in the top-left to open the sidebar
  2. Click the Settings button on the sidebar
  3. Click the Models tab
  4. Click Switch models
  5. Choose from your configured providers in the dropdown, or select Use other provider to configure a new one
  6. Select a model from the available options, or choose Use custom model to enter a specific model name
  7. Click Select model to confirm your choice

:::tip Shortcut For faster access, click your current model name at the bottom of the app and choose Change Model. :::

To start over with provider and model configuration:

  1. Click the button in the top-left to open the sidebar
  2. Click the Settings button on the sidebar
  3. Click the Models tab
  4. Click Reset Provider and Model to clear your current settings and return to the welcome screen
1. In your terminal, run the following command: 
   ```sh
   goose configure
   ```

2. Select `Configure Providers` from the menu and press `Enter`.

   ```
   ┌   goose-configure 
   │
   ◆  What would you like to configure?
   // highlight-start
   │  ● Configure Providers (Change provider or update credentials)
   // highlight-end
   │  ○ Custom Providers 
   │  ○ Add Extension 
   │  ○ Toggle Extensions 
   │  ○ Remove Extension 
   │  ○ goose Settings 
   └  
   ```
3. Choose a model provider and press `Enter`. Use the arrow keys (↑/↓) to move through the options, or start typing to filter the list.

   ```
   ┌   goose-configure 
   │
   ◇  What would you like to configure?
   │  Configure Providers 
   │
   ◆  Which model provider should we use?
   │  ○ Amazon Bedrock 
   │  ○ Amazon SageMaker TGI 
   // highlight-start
   │  ● Anthropic (Claude and other models from Anthropic)
   // highlight-end
   │  ○ Azure OpenAI 
   │  ○ Claude Code CLI
   │  ○ ...
   └  
   ```
4. Enter your API key (and any other configuration details) when prompted.

   ```
   ┌   goose-configure 
   │
   ◇  What would you like to configure?
   │  Configure Providers 
   │
   ◇  Which model provider should we use?
   │  Anthropic 
   │
   ◆  Provider Anthropic requires ANTHROPIC_API_KEY, please enter a value
   // highlight-start
   │  ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
   // highlight-end
   └  
   ```
   
   If you're just changing models, skip any prompts to update the provider configuration.

5. Enter your desired `ANTHROPIC_HOST` or press `Enter` to use the default. 

   ```
   ◆  Provider Anthropic requires ANTHROPIC_HOST, please enter a value
   // highlight-start
   │  https://api.anthropic.com (default)
   // highlight-end
   ```
6. Choose the model you want to use. Depending on the provider, you can:
   - Select the model from a list
   - Search for the model by name
   - Enter the model name directly
   
   ```
   │
   ◇  Model fetch complete
   │
   ◇  Select a model:
   // highlight-start
   │  claude-sonnet-4-5 (default)
   // highlight-end
   │
   ◒  Checking your configuration...
   └  Configuration saved successfully
   ```

   This change takes effect the next time you start a session.

:::note goose configure doesn't support entering custom model names. To use a model not in the provider's list, use goose Desktop or edit the GOOSE_MODEL variable in your config.yaml directly. :::

:::tip Set the model for an individual session using the run command:

goose run --model claude-sonnet-4-0 -t "initial prompt"

:::

Using Custom OpenAI Endpoints

The built-in OpenAI provider can connect to OpenAI's official API (api.openai.com) or any OpenAI-compatible endpoint, such as:

  • Self-hosted LLMs (e.g., LLaMA, Mistral) using vLLM or KServe
  • Private OpenAI-compatible API servers
  • Enterprise deployments requiring data governance and security compliance
  • OpenAI API proxies or gateways

:::tip Custom Provider Option Need to connect to multiple OpenAI-compatible endpoints? Configure custom providers instead for easier switching and better organization, as well as custom naming and shareable configurations. :::

Configuration Parameters

Parameter Required Description
OPENAI_API_KEY Yes Authentication key for the API
OPENAI_HOST No Custom endpoint URL (defaults to api.openai.com)
OPENAI_ORGANIZATION No Organization ID for usage tracking and governance
OPENAI_PROJECT No Project identifier for resource management
OPENAI_CUSTOM_HEADERS No Additional headers to include in the request. Can be set via environment variable, configuration file, or CLI, in the format HEADER_A=VALUE_A,HEADER_B=VALUE_B.

Example Configurations

If you're running LLaMA or other models using vLLM with OpenAI compatibility: ```sh OPENAI_HOST=https://your-vllm-endpoint.internal OPENAI_API_KEY=your-internal-api-key ``` For models deployed on Kubernetes using KServe: ```sh OPENAI_HOST=https://kserve-gateway.your-cluster OPENAI_API_KEY=your-kserve-api-key OPENAI_ORGANIZATION=your-org-id OPENAI_PROJECT=ml-serving ``` For enterprise OpenAI deployments with governance: ```sh OPENAI_API_KEY=your-api-key OPENAI_ORGANIZATION=org-id123 OPENAI_PROJECT=compliance-approved ``` For OpenAI-compatible endpoints that require custom headers: ```sh OPENAI_API_KEY=your-api-key OPENAI_ORGANIZATION=org-id123 OPENAI_PROJECT=compliance-approved OPENAI_CUSTOM_HEADERS="X-Header-A=abc,X-Header-B=def" ```

Setup Instructions

1. Click the button in the top-left to open the sidebar 2. Click the `Settings` button on the sidebar 3. Click the `Models` tab 4. Click `Configure providers` 5. Click `OpenAI` in the provider list 6. Fill in your configuration details: - API Key (required) - Host URL (for custom endpoints) - Organization ID (for usage tracking) - Project (for resource management) 7. Click `Submit` 1. Run `goose configure` 2. Select `Configure Providers` 3. Choose `OpenAI` as the provider 4. Enter your configuration when prompted: - API key - Host URL (if using custom endpoint) - Organization ID (if using organization tracking) - Project identifier (if using project management)

:::tip Enterprise Deployment For enterprise deployments, you can pre-configure these values using environment variables or configuration files to ensure consistent governance across your organization. :::

Configure Custom Provider

Create custom providers to connect to services that aren't already supported or customize how you connect to them. Custom providers appear in goose's provider list and can be selected like any other provider.

Benefits:

  • Multiple endpoints: Switch between different services (e.g., vLLM, corporate proxy, OpenAI)
  • Pre-configured models: Store a list of preferred models
  • Shareable configuration: JSON files can be shared across teams or checked into repos
  • Custom naming: Show "Corporate API" instead of "OpenAI" in the UI
  • Separate credentials: Assign each provider its own API key

Custom providers must use OpenAI, Anthropic, or Ollama compatible API formats. They can include custom headers for additional authentication, API keys, tokens, or tenant identifiers. Each custom provider maps to a JSON configuration file.

To add a custom provider: 1. Click the button in the top-left to open the sidebar 2. Click the Settings button on the sidebar 3. Click the Models tab 4. Click Configure providers 5. Click Add Custom Provider at the bottom of the window 6. Fill in the provider details: - Provider Type: - OpenAI Compatible (most common) - Anthropic Compatible - Ollama Compatible - Display Name: A friendly name for the provider - API URL: The base URL of the API endpoint - Authentication: - API Key: The API key, which is accessed using a custom environment variable and stored in the keychain (or secrets.yaml if the keyring is disabled or cannot be accessed) - For providers that don't require authorization (e.g., local models like Ollama, vLLM, or internal APIs), uncheck the "This provider requires an API key" checkbox - Available Models: Comma-separated list of available model names - Streaming Support: Whether the API supports streaming responses (click to toggle) 7. Click Create Provider

:::info Custom Headers
Currently, custom headers can't be defined in goose Desktop. As a workaround, edit the provider configuration file after creation.
:::
1. In your terminal, run the following command: 
   ```sh
   goose configure
   ```

2. Select `Custom Providers`. Use the arrow keys (↑/↓) to move through the options.

   ```sh
   ┌   goose-configure 
   │
   ◆  What would you like to configure?
   │  ○ Configure Providers
   // highlight-start
   │  ● Custom Providers (Add custom provider with compatible API)
   // highlight-end
   │  ○ Add Extension 
   │  ○ Toggle Extensions 
   │  ○ Remove Extension 
   │  ○ goose Settings 
   └  
   ```

3. Select `Add A Custom Provider`

   ```sh
   ┌   goose-configure 
   │
   ◇  What would you like to configure?
   │  Custom Providers 
   │
   ◆  What would you like to do?
   // highlight-start
   │  ● Add A Custom Provider (Add a new OpenAI/Anthropic/Ollama compatible Provider)
   // highlight-end
   │  ○ Remove Custom Provider
   └  
   ```

4. Follow the prompts to enter the provider details:
   - **API Type**: 
     - `OpenAI Compatible` (most common)
     - `Anthropic Compatible`
     - `Ollama Compatible`
   - **Name**: A friendly name for the provider
   - **API URL**: The base URL of the API endpoint
   - **Authentication Required**: Answer "Yes" if your provider needs an API key, or "No" if authentication is not required
     - If Yes: You'll be prompted to enter your **API Key** (stored securely in the keychain, or in `secrets.yaml` if the keyring is disabled or cannot be accessed)
     - If No: The API key prompt is skipped
   - **Available Models**: Comma-separated list of available model names
   - **Streaming Support**: Whether the API supports streaming responses
   - **Custom Headers**: Any additional header names and values

:::info Custom Headers
Currently, custom headers can only be defined for OpenAI compatible providers in the CLI. For Anthropic or Ollama compatible providers, edit the provider configuration file after creation.
:::

First create a JSON file in the `custom_providers` directory:
- macOS/Linux: `~/.config/goose/custom_providers/`
- Windows: `%APPDATA%\Block\goose\config\custom_providers\`

Example `custom_corp_api.json` configuration file:
```json
{
  "name": "custom_corp_api",
  "engine": "openai",
  "display_name": "Corporate API",
  "description": "Custom Corporate API provider",
  "api_key_env": "CUSTOM_CORP_API_API_KEY",
  "base_url": "https://api.company.com/v1/chat/completions",
  "models": [
    {
      "name": "gpt-4o",
      "context_limit": 128000
    },
    {
      "name": "gpt-3.5-turbo",
      "context_limit": 16385
    }
  ],
  "headers": {
    "x-origin-client-id": "YOUR_CLIENT_ID",
    "x-origin-secret": "YOUR_SECRET_VALUE"
  },
  "supports_streaming": true,
  "requires_auth": true
}
```

Then use the `api_key_env` to set the key for your session. For example:
```bash
export CUSTOM_CORP_API_API_KEY="your-api-key"
goose session start --provider custom_corp_api
```

:::tip Keychain Key Storage
If you want to store the API key in the `goose` keychain, update the provider in goose Desktop and enter the key. This provides secure, persistent storage and allows goose to connect natively to the provider.
:::

To update a custom provider:

1. Click the button in the top-left to open the sidebar 2. Click the `Settings` button on the sidebar 3. Click the `Models` tab 4. Click `Configure providers` 5. Click on your custom provider in the list 6. Update the fields you want to change 7. Click `Update Provider` 
1. In your terminal, run the following command: 

   ```sh
   goose configure
   ```

2. Select `Configure Providers` from the menu and press `Enter`.

   ```sh
   ┌   goose-configure 
   │
   ◆  What would you like to configure?
   // highlight-start
   │  ● Configure Providers (Change provider or update credentials)
   // highlight-end
   │  ○ Custom Providers 
   │  ○ Add Extension 
   │  ○ Toggle Extensions 
   │  ○ Remove Extension 
   │  ○ goose Settings 
   └  
   ```

3. Select the custom provider you want to update and press `Enter`. Use the arrow keys (↑/↓) to move through the options, or start typing to filter the list.

   ```sh
   ┌   goose-configure 
   │
   ◇  What would you like to configure?
   │  Configure Providers 
   │
   ◆  Which model provider should we use?
   │  ○ Amazon Bedrock 
   │  ○ Amazon SageMaker TGI 
   │  ○ Anthropic
   │  ○ Azure OpenAI 
   │  ○ Claude Code CLI 
   // highlight-start
   │  ● Corporate API (Custom Corporate API provider)
   // highlight-end
   │  ○ Cursor Agent 
   │  ○ ...
   └  
   ```

4. Follow the prompts to update the fields.

Open the custom provider configuration file in the `custom_providers` directory:
- macOS/Linux: `~/.config/goose/custom_providers/`
- Windows: `%APPDATA%\Block\goose\config\custom_providers\`

Update the fields you want to change and save your changes.

Your changes are available in your next goose session.

To remove a custom provider:

1. Click the button in the top-left to open the sidebar 2. Click the `Settings` button on the sidebar 3. Click the `Models` tab 4. Click `Configure providers` 5. Click on your custom provider in the list 6. Click `Delete Provider` 7. Confirm that you want to permanently remove the custom provider and its stored API key (if applicable) by clicking `Confirm Delete` 
1. In your terminal, run the following command: 

   ```sh
   goose configure
   ```

2. Select `Custom Providers`. Use the arrow keys (↑/↓) to move through the options.

   ```sh
   ┌   goose-configure 
   │
   ◆  What would you like to configure?
   │  ○ Configure Providers
   // highlight-start
   │  ● Custom Providers (Add custom provider with compatible API)
   // highlight-end
   │  ○ Add Extension 
   │  ○ Toggle Extensions 
   │  ○ Remove Extension 
   │  ○ goose Settings 
   └  
   ```

3. Select `Remove Custom Provider`.

   ```sh
   ┌   goose-configure 
   │
   ◇  What would you like to configure?
   │  Custom Providers 
   │
   ◆  What would you like to do?
   │  ○ Add A Custom Provider 
   // highlight-start
   │  ● Remove Custom Provider (Remove an existing custom provider)
   // highlight-end
   └  
   ```

4. Select the custom provider you want to remove.

The provider configuration file is removed from the `custom_providers` directory and the key is removed from the keychain.

:::tip
If the provider's API key is stored in the keychain, use goose CLI to remove the custom provider. This also removes the stored API key.
:::

Delete the custom provider configuration file in the `custom_providers` directory:
- macOS/Linux: `~/.config/goose/custom_providers/`
- Windows: `%APPDATA%\Block\goose\config\custom_providers\`

Using goose for Free

goose is a free and open source AI agent that you can start using right away, but not all supported LLM Providers provide a free tier.

Below, we outline a couple of free options and how to get started with them.

:::warning Limitations These free options are a great way to get started with goose and explore its capabilities. However, you may need to upgrade your LLM for better performance. :::

Groq

Groq provides free access to open source (open weight) models with high-speed inference. To use Groq with goose, you need an API key from Groq Console.

Groq offers several open source models that support tool calling, including:

  • moonshotai/kimi-k2-instruct-0905 - Mixture-of-Experts model with 1 trillion parameters, optimized for agentic intelligence and tool use
  • qwen/qwen3-32b - 32.8 billion parameter model with advanced reasoning and multilingual capabilities
  • llama-3.3-70b-versatile - Meta's Llama 3.3 model for versatile applications
  • llama-3.1-8b-instant - Meta's Llama 3.1 model for fast inference

For the complete list of supported Groq models, see groq.json.

To set up Groq with goose, follow these steps:

**To update your LLM provider and API key:** 
1. Click the <PanelLeft className="inline" size={16} /> button in the top-left to open the sidebar.
2. Click the `Settings` button on the sidebar.
3. Click the `Models` tab.
4. Click `Configure Providers`
5. Choose `Groq` as provider from the list.
6. Click `Configure`, enter your API key, and click `Submit`.
7. Select the Groq model of your choice.
1. Run: ```sh goose configure ``` 2. Select `Configure Providers` from the menu. 3. Follow the prompts to choose `Groq` as the provider. 4. Enter your API key when prompted. 5. Select the Groq model of your choice.

Google Gemini

Google Gemini provides a free tier. To start using the Gemini API with goose, you need an API Key from Google AI studio.

To set up Google Gemini with goose, follow these steps:

**To update your LLM provider and API key:** 
1. Click the <PanelLeft className="inline" size={16} /> button in the top-left to open the sidebar.
2. Click the `Settings` button on the sidebar.
3. Click the `Models` tab.
4. Click `Configure Providers`
5. Choose `Google Gemini` as provider from the list.
6. Click `Configure`, enter your API key, and click `Submit`.
1. Run: ```sh goose configure ``` 2. Select `Configure Providers` from the menu. 3. Follow the prompts to choose `Google Gemini` as the provider. 4. Enter your API key when prompted. 5. Enter the Gemini model of your choice. 
```
┌   goose-configure
│
◇ What would you like to configure?
│ Configure Providers
│
◇ Which model provider should we use?
│ Google Gemini
│
◇ Provider Google Gemini requires GOOGLE_API_KEY, please enter a value
│▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
│    
◇ Enter a model from that provider:
│ gemini-2.0-flash-exp
│
◇ Hello! You're all set and ready to go, feel free to ask me anything!
│
└ Configuration saved successfully
```

Local LLMs

goose is a local AI agent, and by using a local LLM, you keep your data private, maintain full control over your environment, and can work entirely offline without relying on cloud access. However, please note that local LLMs require a bit more set up before you can use one of them with goose.

:::warning Limited Support for models without tool calling goose extensively uses tool calling, so models without it can only do chat completion. If using models without tool calling, all goose extensions must be disabled. :::

Here are some local providers we support:

1. [Download Ramalama](https://github.com/containers/ramalama?tab=readme-ov-file#install). 2. In a terminal, run any Ollama [model supporting tool-calling](https://ollama.com/search?c=tools) or [GGUF format HuggingFace Model](https://huggingface.co/search/full-text?q=%22tools+support%22+%2B+%22gguf%22&type=model): 
      The `--runtime-args="--jinja"` flag is required for Ramalama to work with the goose Ollama provider.

      Example:

      ```sh
      ramalama serve --runtime-args="--jinja" ollama://qwen2.5
      ```

      3. In a separate terminal window, configure with goose:

      ```sh
      goose configure
      ```

      4. Choose to `Configure Providers`

      ```
      ┌   goose-configure
      │
      ◆  What would you like to configure?
      │  ● Configure Providers (Change provider or update credentials)
      │  ○ Toggle Extensions
      │  ○ Add Extension
      └
      ```

      5. Choose `Ollama` as the model provider since Ramalama is API compatible and can use the goose Ollama provider

      ```
      ┌   goose-configure
      │
      ◇  What would you like to configure?
      │  Configure Providers
      │
      ◆  Which model provider should we use?
      │  ○ Anthropic
      │  ○ Databricks
      │  ○ Google Gemini
      │  ○ Groq
      │  ● Ollama (Local open source models)
      │  ○ OpenAI
      │  ○ OpenRouter
      └
      ```

      6. Enter the host where your model is running

      :::info Endpoint
      For the Ollama provider, if you don't provide a host, we set it to `localhost:11434`. When constructing the URL, we prepend `http://` if the scheme is not `http` or `https`. Since Ramalama's default port to serve on is 8080, we set `OLLAMA_HOST=http://0.0.0.0:8080`
      :::

      ```
      ┌   goose-configure
      │
      ◇  What would you like to configure?
      │  Configure Providers
      │
      ◇  Which model provider should we use?
      │  Ollama
      │
      ◆  Provider Ollama requires OLLAMA_HOST, please enter a value
      │  http://0.0.0.0:8080
      └
      ```


      7. Enter the model you have running

      ```
      ┌   goose-configure
      │
      ◇  What would you like to configure?
      │  Configure Providers
      │
      ◇  Which model provider should we use?
      │  Ollama
      │
      ◇  Provider Ollama requires OLLAMA_HOST, please enter a value
      │  http://0.0.0.0:8080
      │
      ◇  Enter a model from that provider:
      │  qwen2.5
      │
      ◇  Welcome! You're all set to explore and utilize my capabilities. Let's get started on solving your problems together!
      │
      └  Configuration saved successfully
      ```

      :::tip Context Length
      If you notice that goose is having trouble using extensions or is ignoring [.goosehints](/docs/guides/context-engineering/using-goosehints), it is likely that the model's default context length of 2048 tokens is too low. Use `ramalama serve` to set the `--ctx-size, -c` option to a [higher value](https://github.com/containers/ramalama/blob/main/docs/ramalama-serve.1.md#--ctx-size--c).
      :::

  </TabItem>
  <TabItem value="deepseek" label="DeepSeek-R1">
    The native `DeepSeek-r1` model doesn't support tool calling, however, we have a [custom model](https://ollama.com/michaelneale/deepseek-r1-goose) you can use with goose. 

    :::warning
    Note that this is a 70B model size and requires a powerful device to run smoothly.
    :::


    1. [Download Ollama](https://ollama.com/download). 
    2. In a terminal window, run the following command to install the custom DeepSeek-r1 model:

    ```sh
    ollama run michaelneale/deepseek-r1-goose
    ```

    3. In a separate terminal window, configure with goose:

    ```sh
    goose configure
    ```

    4. Choose to `Configure Providers`

    ```
    ┌   goose-configure 
    │
    ◆  What would you like to configure?
    │  ● Configure Providers (Change provider or update credentials)
    │  ○ Toggle Extensions 
    │  ○ Add Extension 
    └  
    ```

    5. Choose `Ollama` as the model provider

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◆  Which model provider should we use?
    │  ○ Anthropic 
    │  ○ Databricks 
    │  ○ Google Gemini 
    │  ○ Groq 
    │  ● Ollama (Local open source models)
    │  ○ OpenAI 
    │  ○ OpenRouter 
    └  
    ```

    6. Enter the host where your model is running

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◇  Which model provider should we use?
    │  Ollama 
    │
    ◆  Provider Ollama requires OLLAMA_HOST, please enter a value
    │  http://localhost:11434
    └
    ```

    7. Enter the installed model from above

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◇  Which model provider should we use?
    │  Ollama 
    │
    ◇   Provider Ollama requires OLLAMA_HOST, please enter a value
    │  http://localhost:11434  
    │    
    ◇  Enter a model from that provider:
    │  michaelneale/deepseek-r1-goose
    │
    ◇  Welcome! You're all set to explore and utilize my capabilities. Let's get started on solving your problems together!
    │
    └  Configuration saved successfully
    ```
  </TabItem>
  <TabItem value="others" label="Other Models" default>
    1. [Download Ollama](https://ollama.com/download). 
    2. In a terminal, run any [model supporting tool-calling](https://ollama.com/search?c=tools)

      Example:

      ```sh
      ollama run qwen2.5
      ```

    3. In a separate terminal window, configure with goose:

      ```sh
      goose configure
      ```

    4. Choose to `Configure Providers`

    ```
    ┌   goose-configure 
    │
    ◆  What would you like to configure?
    │  ● Configure Providers (Change provider or update credentials)
    │  ○ Toggle Extensions 
    │  ○ Add Extension 
    └  
    ```

    5. Choose `Ollama` as the model provider

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◆  Which model provider should we use?
    │  ○ Anthropic 
    │  ○ Databricks 
    │  ○ Google Gemini 
    │  ○ Groq 
    │  ● Ollama (Local open source models)
    │  ○ OpenAI 
    │  ○ OpenRouter 
    └  
    ```

    6. Enter the host where your model is running

    :::info Endpoint
    For Ollama, if you don't provide a host, we set it to `localhost:11434`. 
    When constructing the URL, we prepend `http://` if the scheme is not `http` or `https`. 
    If you're running Ollama on a different server, you'll have to set `OLLAMA_HOST=http://{host}:{port}`.
    :::

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◇  Which model provider should we use?
    │  Ollama 
    │
    ◆  Provider Ollama requires OLLAMA_HOST, please enter a value
    │  http://localhost:11434
    └
    ```


    7. Enter the model you have running

    ```
    ┌   goose-configure 
    │
    ◇  What would you like to configure?
    │  Configure Providers 
    │
    ◇  Which model provider should we use?
    │  Ollama 
    │
    ◇  Provider Ollama requires OLLAMA_HOST, please enter a value
    │  http://localhost:11434
    │
    ◇  Enter a model from that provider:
    │  qwen2.5
    │
    ◇  Welcome! You're all set to explore and utilize my capabilities. Let's get started on solving your problems together!
    │
    └  Configuration saved successfully
    ```

    :::tip Context Length
    If you notice that goose is having trouble using extensions or is ignoring [.goosehints](/docs/guides/context-engineering/using-goosehints), it is likely that the model's default context length of 4096 tokens is too low. Set the `OLLAMA_CONTEXT_LENGTH` environment variable to a [higher value](https://github.com/ollama/ollama/blob/main/docs/faq.mdx#how-can-i-specify-the-context-window-size).
    :::
    
  </TabItem>
</Tabs>
[LM Studio](https://lmstudio.ai/) lets you run open-source models locally with an OpenAI-compatible API server. 
1. Download and install LM Studio.
2. Open LM Studio and download a model that supports tool calling (e.g., Qwen, Llama, or Mistral variants).
3. Start the local server in LM Studio. The server runs on `http://localhost:1234` by default

4. Configure goose to use LM Studio:

<Tabs groupId="interface">
  <TabItem value="ui" label="goose Desktop" default>
    1. Click the <PanelLeft className="inline" size={16} /> button in the top-left to open the sidebar.
    2. Click the `Settings` button on the sidebar.
    3. Click the `Models` tab.
    4. Click `Configure providers`.
    5. Choose `LM Studio` from the provider list and click `Configure`.
    6. Click `Submit` (no API key is needed).
    7. Select the model you have loaded in LM Studio.
  </TabItem>
  <TabItem value="cli" label="goose CLI">
    1. Run:
    ```sh
    goose configure
    ```
    2. Select `Configure Providers` from the menu.
    3. Choose `LM Studio` as the provider.
    4. Enter the model name that matches the model loaded in LM Studio.

    ```
    ┌   goose-configure
    │
    ◇  What would you like to configure?
    │  Configure Providers
    │
    ◇  Which model provider should we use?
    │  LM Studio
    │
    ◇  Enter a model from that provider:
    │  qwen2.5-7b-instruct
    │
    └  Configuration saved successfully
    ```
  </TabItem>
</Tabs>

:::tip Model Name
Make sure the model name you enter in goose matches the model identifier shown in LM Studio's server panel.
:::
1. [Get Docker](https://docs.docker.com/get-started/get-docker/) 2. [Enable Docker Model Runner](https://docs.docker.com/ai/model-runner/#enable-dmr-in-docker-desktop) 3. [Pull a model](https://docs.docker.com/ai/model-runner/#pull-a-model), for example, from Docker Hub [AI namespace](https://hub.docker.com/u/ai), [Unsloth](https://hub.docker.com/u/unsloth), or [from HuggingFace](https://www.docker.com/blog/docker-model-runner-on-hugging-face/) 
Example:

```sh
docker model pull hf.co/unsloth/gemma-3n-e4b-it-gguf:q6_k
```

4. Configure goose to use Docker Model Runner, using the OpenAI API compatible endpoint: 

```sh
goose configure
```

5. Choose to `Configure Providers`

```
┌   goose-configure 
│
◆  What would you like to configure?
│  ● Configure Providers (Change provider or update credentials)
│  ○ Toggle Extensions 
│  ○ Add Extension 
└  
```

6. Choose `OpenAI` as the model provider: 

```
┌   goose-configure
│
◇  What would you like to configure?
│  Configure Providers
│
◆  Which model provider should we use?
│  ○ Anthropic
│  ○ Amazon Bedrock
│  ○ Claude Code
│  ● OpenAI (GPT-4 and other OpenAI models, including OpenAI compatible ones)
│  ○ OpenRouter
```

7. Configure Docker Model Runner endpoint as the `OPENAI_HOST`: 

```
┌   goose-configure
│
◇  What would you like to configure?
│  Configure Providers
│
◇  Which model provider should we use?
│  OpenAI
│
◆  Provider OpenAI requires OPENAI_HOST, please enter a value
│  https://api.openai.com (default)
└
```

The default value for the host-side port Docker Model Runner is 12434, so the `OPENAI_HOST` value could be: 
`http://localhost:12434`. 

8. Configure the base path: 

```
◆  Provider OpenAI requires OPENAI_BASE_PATH, please enter a value
│  v1/chat/completions (default)
└
```

Docker model runner uses `/engines/llama.cpp/v1/chat/completions` for the base path.

9. Finally configure the model available in Docker Model Runner to be used by goose: `hf.co/unsloth/gemma-3n-e4b-it-gguf:q6_k`

```
│
◇  Enter a model from that provider:
│  gpt-4o
│
◒  Checking your configuration...                                                                                                            
└  Configuration saved successfully
```

GitHub Copilot Authentication

GitHub Copilot uses a device flow for authentication, so no API keys are required:

  1. Run goose configure and select GitHub Copilot
  2. An eight-character code will be automatically copied to your clipboard
  3. A browser will open to GitHub's device activation page
  4. Paste the code to authorize the application
  5. When you return to goose, GitHub Copilot will be available as a provider in both CLI and Desktop.

Azure OpenAI Credential Chain

goose supports two authentication methods for Azure OpenAI:

  1. API Key Authentication - Uses the AZURE_OPENAI_API_KEY for direct authentication
  2. Azure Credential Chain - Uses Azure CLI credentials automatically without requiring an API key

To use the Azure Credential Chain:

  • Ensure you're logged in with az login
  • Have appropriate Azure role assignments for the Azure OpenAI service
  • Configure with goose configure and select Azure OpenAI, leaving the API key field empty

This method simplifies authentication and enhances security for enterprise environments.

Multi-Model Configuration

Beyond single-model setups, goose supports multi-model configurations that can use different models and providers for specialized tasks:

  • Planning Mode - Use a dedicated planner model to create detailed project breakdowns before execution
  • Subagents - Delegate scoped tasks to isolated sessions to keep your primary workflow focused and efficient

Gemini 3 Thinking Levels

Gemini 3 models support configurable thinking levels to balance response latency and reasoning depth:

  • Low (default) - Faster responses, lighter reasoning
  • High - Deeper reasoning, higher latency

:::tip When thinking is enabled, you can view the model's reasoning process. See Viewing Model Reasoning for details. :::

When selecting a Gemini 3 model, a "Thinking Level" dropdown appears automatically. Select your preference and the setting persists across sessions. **Interactive configuration:** 
When you run `goose configure` and select a Gemini 3 model, you'll be prompted to choose a thinking level:

```
◆  Select thinking level for Gemini 3:
│  ● Low - Better latency, lighter reasoning
│  ○ High - Deeper reasoning, higher latency
```

:::info Priority Order The thinking level is determined in this order (highest to lowest priority):

  1. request_params.thinking_level in model configuration (via GOOSE_PREDEFINED_MODELS)
  2. GEMINI3_THINKING_LEVEL environment variable
  3. Default value: low :::

Viewing Model Reasoning

Some models expose their internal reasoning or "chain of thought" as part of their response. goose automatically captures this reasoning output and makes it available to you. The following models and providers support reasoning output:

Provider / Model How It Works
DeepSeek-R1 (via OpenAI, Ollama, OpenRouter, OVHcloud, etc.) Reasoning captured from the reasoning_content field in the API response
Kimi (via Groq or other OpenAI-compatible endpoints) Reasoning captured from the reasoning_content field in the API response
Gemini CLI (Google Gemini models with thinking enabled) Thinking blocks captured from the streaming response
Claude (Anthropic, with Claude thinking enabled) Thinking blocks captured from the API response
Reasoning output appears automatically in a collapsible **"Show reasoning"** toggle above the model's response. Click it to expand and view the model's thought process. Reasoning output is **hidden by default** in the CLI. To display it, set the `GOOSE_CLI_SHOW_THINKING` environment variable: 
```bash
export GOOSE_CLI_SHOW_THINKING=1
```

When enabled, reasoning appears under a "Thinking:" header in dimmed text before the model's main response.

:::note
This requires stdout to be a terminal (reasoning output won't appear when piping output to a file or another command).
:::

:::tip Reasoning output can be useful for understanding how the model arrived at its answer, debugging unexpected behavior, or learning from the model's problem-solving approach. However, it can also be verbose — toggle it on only when you need it. :::

If you have any questions or need help with a specific provider, feel free to reach out to us on Discord or on the goose repo.