Tools, Editors & CLI Agents

Overview

Bifrost provides 100% compatible endpoints for OpenAI, Anthropic, and Gemini APIs, making it seamless to integrate with any agent that uses these providers. By simply pointing your agent’s base URL to Bifrost, you unlock powerful features like:

Universal Model Access: Use any provider/model configured in Bifrost with any agent (e.g., use GPT-5 with Claude Code, or Claude Sonnet 4.5 with Codex CLI)
MCP Tools Integration: All Model Context Protocol tools configured in Bifrost become available to your agents
Built-in Observability: Monitor all agent interactions in real-time through Bifrost’s logging dashboard
Load Balancing: Automatically distribute requests across multiple providers and regions
Advanced Features: Governance, caching, failover, and more - all transparent to your agent

Example Integrations

LibreChat

It is a modern, open-source chat client that supports multiple providers. Setup:

Install LibreChat: There are multiple ways of local setup, please follow the LibreChat documentation for more details.

Add Bifrost as a custom provider: Now that you have LibreChat installed, you can add Bifrost as a custom provider. Add the following to your librechat.yaml file:

custom:
 - name: "Bifrost"
   apiKey: "dummy" # Add the authentication key if login is enabled, otherwise add a placeholder
   baseURL: "http://host.docker.internal:8080/v1" # Or localhost:8080 if running locally, or {your-bifrost-container}:8080 if running in the same docker network
   models:
     default: ["openai/gpt-4o"] # Replace with the model you want to use
     fetch: true
   titleConvo: true
   titleModel: "openai/gpt-4o" # Replace with the model you want to use for chat title generation
   summarize: false # Set to true if you want to enable chat summary generation
   summaryModel: "openai/gpt-4o" # Replace with the model you want to use for chat summary generation
   forcePrompt: false # Set to true if you want to enable force prompt generation
   modelDisplayLabel: "Bifrost" 
   iconURL: https://getbifrost.ai/bifrost-logo.png

If you’re running LibreChat in a docker container, LibreChat does not automatically use the librechat.yaml file, please check the Step 1 of the LibreChat documentation for more details.

Run LibreChat Now you can start using Bifrost as a provider in LibreChat, with all the features of Bifrost.

Claude Code

It brings Sonnet 4.5 directly to your terminal with powerful coding capabilities. Setup:

Install Claude Code

npm install -g @anthropic-ai/claude-code

Configure Environment Variables

export ANTHROPIC_API_KEY=dummy-key  # Handled by Bifrost (only set when using virtual keys)
export ANTHROPIC_BASE_URL=http://localhost:8080/anthropic

Run Claude Code
```
claude
```

Now all Claude Code traffic flows through Bifrost, giving you access to any provider/model configured in your Bifrost setup, plus MCP tools and observability.

This setup automatically detects if you’re using Anthropic MAX account instead of a regular API key authentication :)

Codex CLI

It provides powerful code generation and completion capabilities. Setup:

Install Codex CLI
```
npm install -g @openai/codex
```

Configure Environment Variables

export OPENAI_API_KEY=dummy-key  # Handled by Bifrost (only set when using virtual keys)
export OPENAI_BASE_URL=http://localhost:8080/openai

Run Codex
```
codex
```

Gemini CLI

It is Google’s powerful coding assistant with advanced reasoning capabilities. Setup:

Install Gemini CLI
```
npm install -g @google/gemini-cli
```

Configure Environment Variables

export GEMINI_API_KEY=dummy-key  # Handled by Bifrost (only set when using virtual keys)
export GOOGLE_GEMINI_BASE_URL=http://localhost:8080/genai

Run Gemini CLI
```
gemini
```
Select Use Gemini API Key in the CLI prompt for authentication.

Qwen Code

It is Alibaba’s powerful coding assistant with advanced reasoning capabilities. Setup:

Install Qwen Code
```
 npm install -g @qwen-code/qwen-code
```

Configure Base URL

export OPENAI_BASE_URL=http://localhost:8080/openai

Run Qwen Code
```
qwen
```

Opencode

Setup

Configure Bifrost

{
  "$schema": "https://opencode.ai/config.json",
  // Theme configuration
  "theme": "opencode",
  "autoupdate": true,
  "provider": {
    "openai": {
      "name": "Bifrost",
      "options": {        
        "baseURL": "http://localhost:8080/openai",
        "apiKey": "{{virtual-key-if-enabled}}"
      },
      "models": {
        "openai/gpt-5": {
          "options": {
            "reasoningEffort": "high",
            "textVerbosity": "low",
            "reasoningSummary": "auto",
            "include": [
              "reasoning.encrypted_content"
            ],
          },
        },
        "anthropic/claude-sonnet-4-5-20250929": {
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 16000,
            },
          },
        },
      },
    }
  }
}

Select Bifrost models using ctrl+p

Editors

Zed editor

Configure Bifrost provider.

   "language_models": {
        "openai_compatible": {
            "Bifrost": {
                "api_url": "{{bifrost-base-url}}/openai",
                "available_models": [
                    {
                        "name": "anthropic/claude-sonnet-4.5",
                        "max_tokens": 200000,
                        "max_output_tokens": 4096,
                        "capabilities": {
                            "tools": true,
                            "images": true,
                            "parallel_tool_calls": true,
                            "prompt_cache_key": false
                        }
                    },
                    {
                        "name": "openai/gpt-4o",
                        "max_tokens": 128000,
                        "max_output_tokens": 4096,
                        "capabilities": {
                            "tools": true,
                            "images": true,
                            "parallel_tool_calls": true,
                            "prompt_cache_key": false
                        }
                    },
                    {
                        "name": "openai/gpt-5",
                        "max_tokens": 256000,
                        "max_output_tokens": 4096,
                        "capabilities": {
                            "tools": true,
                            "images": true,
                            "parallel_tool_calls": true,
                            "prompt_cache_key": false
                        }
                    }
                ]
            }
        }
    }

Reload workspace to make sure Zed editor recognizes and reloads the provider list.

Configuration

Agent integrations work with your existing Bifrost configuration. Ensure you have:

Providers configured: See Provider Configuration for setup details
Optional: MCP tools: See MCP Integration to enhance agent capabilities

Monitoring Agent Traffic

All agent interactions are automatically logged and can be monitored at http://localhost:8080/logs. You can filter by provider, model, or search through conversation content to track your agents’ performance.

For complete monitoring capabilities, see Built-in Observability.

MCP Tools Integration

Bifrost automatically sends all configured MCP tools to your agents. This means your agents can access filesystem operations, database queries, web search, and more without any additional configuration.

Important: MCP Tool Execution BehaviorWhen using Bifrost as a gateway, MCP tool calls require manual approval and execution for security reasons. Bifrost returns the tool call information but doesn’t automatically execute it. You need to handle the approval and execution logic by calling the v1/mcp/tool/execute endpoint.Gateway-on-Gateway Limitations: If your agent/editor (like Zed) has its own gateway that routes through Bifrost, the agent’s gateway may not handle MCP tool approvals that come from Bifrost. In such cases, we recommend configuring MCP tools directly in your agent/editor instead of relying on Bifrost’s MCP integration.We intentionally avoid supporting “gateway-on-gateway” MCP setups because handling tool approvals across multiple gateways introduces unnecessary complexity and falls outside the scope of what an LLM gateway should manage. While we’re working on an agentic mode that will allow Bifrost to automatically execute certain tool calls, the current design prioritizes security and clear responsibility boundaries.

For setup and available tools, see MCP Integration.

Next Steps

Provider Configuration - Configure AI providers for your agents
Governance - Set usage limits and policies for your agents
Integrations - Understand how Bifrost works with existing AI provider SDKs

Quick Start

Models Catalog

Provider Integrations

Custom plugins

Open Source Features

Enterprise Features

Tools, Editors & CLI Agents

Overview

Example Integrations

LibreChat

Claude Code

Codex CLI

Gemini CLI

Qwen Code

Opencode

Editors

Zed editor

Configuration

Monitoring Agent Traffic

MCP Tools Integration

Next Steps

Quick Start

Models Catalog

Provider Integrations

Custom plugins

Open Source Features

Enterprise Features

​Overview

​Example Integrations

​LibreChat

​Claude Code

​Codex CLI

​Gemini CLI

​Qwen Code

​Opencode

​Editors

​Zed editor

​Configuration

​Monitoring Agent Traffic

​MCP Tools Integration

​Next Steps

Overview

Example Integrations

LibreChat

Claude Code

Codex CLI

Gemini CLI

Qwen Code

Opencode

Editors

Zed editor

Configuration

Monitoring Agent Traffic

MCP Tools Integration

Next Steps