getmaxim.ai

# Bifrost

> Bifrost is an open-source, high-performance AI gateway from Maxim. It puts a single OpenAI-compatible API in front of 1000+ models with automatic failover, load balancing, semantic caching, governance, guardrails, MCP support, and full observability.

## Product

- [LLM gateway](https://www.getmaxim.ai/llm-gateway): One OpenAI-compatible API for 1000+ models with automatic failover, load balancing, and semantic caching.
- [MCP gateway](https://www.getmaxim.ai/mcp-gateway): Connect, govern, and observe Model Context Protocol servers and tools from a single control plane.
- [AI governance](https://www.getmaxim.ai/ai-governance): Virtual keys, budgets, rate limits, and access policies across teams, customers, and environments.
- [AI guardrails](https://www.getmaxim.ai/ai-guardrails): Input and output guardrails for PII, jailbreaks, and unsafe content.
- [AI observability](https://www.getmaxim.ai/ai-observability): Traces, logs, and metrics for every request routed through the gateway.
- [Bifrost Edge](https://www.getmaxim.ai/edge): Run the gateway at the edge, close to your users.
- [Pricing](https://www.getmaxim.ai/pricing): Plans, limits, and what is included in the open-source gateway. (markdown: https://www.getmaxim.ai/pricing.md)
- [Book a Demo | Bifrost Enterprise AI Gateway](https://www.getmaxim.ai/book-a-demo): The fastest, most resilient, enterprise-grade LLM, MCP, and agent gateway. Routing, governance, guardrails, and observability via one control plane. Book a demo.

## Docs and guides

- [Documentation](https://www.getmaxim.ai/bifrost/docs): Installation, configuration, provider setup, and API reference.
- [Resources](https://www.getmaxim.ai/bifrost/resources): Benchmarks, migration guides, buyer guides, and enterprise deployment material. (markdown: https://www.getmaxim.ai/bifrost/resources.md)
- [Bifrost benchmark results](https://www.getmaxim.ai/bifrost/resources/benchmarks): Performance measurements for Bifrost and LiteLLM under sustained load, gateway overhead tests, and high-throughput stress tests. (markdown: https://www.getmaxim.ai/bifrost/resources/benchmarks.md)
- [Bifrost LiteLLM alternative](https://www.getmaxim.ai/bifrost/resources/litellm-alternative): LLM gateway comparison between Bifrost and LiteLLM. Covers OpenAI-compatible routing, Go vs Python gateway architecture, setup steps, observability, governance, provider coverage, and published benchmark context. (markdown: https://www.getmaxim.ai/bifrost/resources/litellm-alternative.md)
- [Migrating from LiteLLM to Bifrost](https://www.getmaxim.ai/bifrost/resources/migrating-from-litellm): Migration guide for moving LiteLLM-compatible traffic to Bifrost. Covers base URL changes, provider setup, virtual key migration, and benchmark context from the 500 RPS LiteLLM comparison. (markdown: https://www.getmaxim.ai/bifrost/resources/migrating-from-litellm.md)
- [LLM gateway buyer's guide](https://www.getmaxim.ai/bifrost/resources/buyers-guide): Evaluation guide for LLM gateways covering deployment model, data control, routing, observability, governance, cost controls, and Bifrost benchmark context. (markdown: https://www.getmaxim.ai/bifrost/resources/buyers-guide.md)
- [Bifrost OSS for Startups | Free Production-Ready LLM Gateway](https://www.getmaxim.ai/bifrost/resources/oss-for-startups): Free, open-source LLM gateway with automatic failover, load balancing, semantic caching, and 20+ provider support. Production-ready features for startups and SMEs at zero cost. (markdown: https://www.getmaxim.ai/bifrost/resources/oss-for-startups.md)
- [Claude Code + Bifrost | Enterprise LLM Gateway for Claude Code](https://www.getmaxim.ai/bifrost/resources/claude-code): Add multi-provider routing, cost control, guardrails, and governance to Claude Code at scale with Bifrost - the fastest enterprise LLM gateway. (markdown: https://www.getmaxim.ai/bifrost/resources/claude-code.md)
- [CLI Coding Agents + Bifrost | Claude Code, Codex, Gemini, OpenCode](https://www.getmaxim.ai/bifrost/resources/cli-agents): Run Claude Code, Codex CLI, Gemini CLI, and OpenCode through Bifrost for multi-provider routing, enterprise controls, cost tracking, and centralized observability. (markdown: https://www.getmaxim.ai/bifrost/resources/cli-agents.md)
- [Bifrost CLI | Launch Claude Code & Coding Agents Through Bifrost](https://www.getmaxim.ai/bifrost/resources/bifrost-cli): Interactive terminal tool to launch Claude Code, Codex CLI, Gemini CLI, and Opencode through Bifrost gateway with automatic configuration, MCP integration, and secure credential storage. (markdown: https://www.getmaxim.ai/bifrost/resources/bifrost-cli.md)
- [Bifrost for AWS Bedrock | Enterprise AI Gateway with Governance & Guardrails](https://www.getmaxim.ai/bifrost/resources/aws-bedrock): Add governance, guardrails, multi-region failover, and cost controls to AWS Bedrock. Bifrost is the enterprise AI gateway with native Bedrock SDK compatibility. Zero code changes. (markdown: https://www.getmaxim.ai/bifrost/resources/aws-bedrock.md)
- [MCP Gateway | High-Performance Tool Execution for AI Agents](https://www.getmaxim.ai/bifrost/resources/mcp-gateway): Connect AI models to external tools with an open-source MCP gateway that delivers 11µs overhead and complete security control. (markdown: https://www.getmaxim.ai/bifrost/resources/mcp-gateway.md)
- [Bifrost Governance | Virtual Keys, Budgets & Enterprise RBAC](https://www.getmaxim.ai/bifrost/resources/governance): Comprehensive governance for LLM access with virtual keys, budgets, rate limits, routing, MCP tool filtering, and enterprise RBAC with SSO integration. (markdown: https://www.getmaxim.ai/bifrost/resources/governance.md)
- [Bifrost Guardrails | Enterprise AI Safety & Policy Enforcement](https://www.getmaxim.ai/bifrost/resources/guardrails): Real-time validation of LLM inputs and outputs with PII detection, content moderation, prompt injection defense, and multi-provider guardrail support for enterprise compliance. (markdown: https://www.getmaxim.ai/bifrost/resources/guardrails.md)
- [Enterprise AI Gateway for Scalability](https://www.getmaxim.ai/bifrost/resources/enterprise-scalability): Scale AI workloads with Bifrost routing, circuit breakers, semantic caching, cost analytics, RBAC, guardrails, in-VPC deployment, and MCP gateway controls for production agents. (markdown: https://www.getmaxim.ai/bifrost/resources/enterprise-scalability.md)
- [Enterprise AI Gateway Deployment | In-VPC, Air-Gapped, Multi-Cloud](https://www.getmaxim.ai/bifrost/resources/enterprise-deployment): Deployment guide for running Bifrost in a VPC, on-premise, air-gapped, or multi-cloud environment with zero data egress and enterprise compliance controls. (markdown: https://www.getmaxim.ai/bifrost/resources/enterprise-deployment.md)

## Guides

- [Provider guides](https://www.getmaxim.ai/bifrost/guides/providers): Setup, supported operations, model prefixes, and endpoint mappings for every provider Bifrost supports.
- [OpenAI on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/openai): Use OpenAI models through Bifrost with full API compatibility. Learn supported operations, endpoint mappings, models, reasoning support, and implementation patterns.
- [Anthropic on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/anthropic): Use Anthropic models through Bifrost with OpenAI-compatible requests. Learn supported operations, endpoint mappings, beta headers, reasoning conversion, and production setup patterns.
- [Azure on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/azure): Use Azure OpenAI Service models through Bifrost with automatic deployment mapping. Learn supported operations, authentication methods, and multi-model routing.
- [AWS Bedrock on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/bedrock): Use AWS Bedrock models through Bifrost with OpenAI-compatible requests. Learn supported operations, model families, authentication, and parameter mappings.
- [Google Gemini on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/gemini): Use Google Gemini models through Bifrost with OpenAI-compatible requests. Learn supported operations, vision support, and parameter mappings.
- [Google Vertex AI on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/vertex-ai): Use Google Vertex AI models through Bifrost with full API compatibility. Learn supported operations, endpoint mappings, models, authentication, and parameter handling.
- [Cohere on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/cohere): Use Cohere models through Bifrost with OpenAI-compatible requests. Learn supported operations, endpoint mappings, embeddings, and reranking capabilities.
- [Mistral on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/mistral): Use Mistral models through Bifrost with OpenAI-compatible requests. Learn supported operations, tool calling limitations, and transcription capabilities.
- [Groq on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/groq): Use Groq models through Bifrost with OpenAI-compatible requests. Learn supported operations, fast inference, and tool calling support.
- [OpenRouter on Bifrost: Multi-Provider LLM Routing, Models, and API Reference](https://www.getmaxim.ai/bifrost/guides/providers/openrouter): Use OpenRouter on Bifrost to route requests across 100+ models from OpenAI, Anthropic, Google, Meta, and more. Learn operations, model IDs, parameter mapping, and setup.
- [Fireworks AI on Bifrost: Models, Endpoints, Streaming, and API Reference](https://www.getmaxim.ai/bifrost/guides/providers/fireworks): Use Fireworks AI models through Bifrost with native Responses API support. Learn supported operations, endpoint mappings, models, streaming, and parameter handling.
- [Cerebras on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/cerebras): Use Cerebras models through Bifrost with OpenAI-compatible requests. Learn supported operations, fast inference, and open-source model routing.
- [Hugging Face on Bifrost: Multi-Backend LLM, Audio, and Image Operations](https://www.getmaxim.ai/bifrost/guides/providers/huggingface): Use Hugging Face inference providers through Bifrost with support for chat, embeddings, TTS, STT, and image generation. Learn operations, models, and setup.
- [Ollama on Bifrost: Self-Hosted LLM Inference, OpenAI Compatibility](https://www.getmaxim.ai/bifrost/guides/providers/ollama): Use Ollama with Bifrost for self-hosted LLM inference. Full OpenAI API compatibility with streaming, embeddings, and tool calling support.
- [vLLM on Bifrost: OpenAI-Compatible Inference, Reranking, and Streaming](https://www.getmaxim.ai/bifrost/guides/providers/vllm): Use vLLM for self-hosted LLM inference with Bifrost. OpenAI-compatible chat, text completions, embeddings, rerank, STT, and streaming.
- [SGLang on Bifrost: OpenAI-Compatible Inference and API Reference](https://www.getmaxim.ai/bifrost/guides/providers/sglang): Use SGLang (SGL) for local or remote LLM inference with Bifrost. OpenAI-compatible API with streaming, embeddings, tool calling, and parameter filtering.
- [xAI Grok on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/xai): Use xAI Grok models through Bifrost with OpenAI-compatible requests. Learn supported operations, vision capabilities, and reasoning support.
- [ElevenLabs on Bifrost: Text-to-Speech and Transcription APIs](https://www.getmaxim.ai/bifrost/guides/providers/elevenlabs): Use ElevenLabs speech and audio services through Bifrost. TTS with voice control, transcription with diarization, and advanced audio features.
- [Replicate on Bifrost: Models, Predictions, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/replicate): Use Replicate through Bifrost with prediction-based chat, images, video, and files. Learn model IDs, sync/async modes, and extra_params.
- [Runway ML on Bifrost: Video Generation, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/runway): Use Runway ML through Bifrost for text-to-video, image-to-video, and video-to-video generation with async task polling.
- [Nebius on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/nebius): Use Nebius through Bifrost with OpenAI-compatible chat, embeddings, image generation, and streaming. Learn supported operations, ai_project_id, and parameter handling.
- [Perplexity on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/perplexity): Use Perplexity through Bifrost with OpenAI-compatible chat, built-in web search, citations, and reasoning support.
- [Parasail on Bifrost: Models, Endpoints, Setup, and Mappings](https://www.getmaxim.ai/bifrost/guides/providers/parasail): Use Parasail through Bifrost for high-performance OpenAI-compatible inference with streaming, tool calling, and reasoning support.
- [API key guides](https://www.getmaxim.ai/bifrost/guides/api-keys): Step-by-step guides for creating, configuring, and routing provider API keys through Bifrost.
- [How to Get an Anthropic API Key (Claude)](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-anthropic-api-key): Sign up at console.anthropic.com, add billing, generate your Claude API key (sk-ant- format), and integrate with Bifrost for cost tracking and virtual key support. 5-minute setup.
- [How to Get an AWS Bedrock API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-aws-bedrock-api-key): Set up AWS Bedrock access through IAM, configure credentials, and integrate with Bifrost for cost tracking and multi-model governance.
- [How to Get an Azure OpenAI API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-azure-openai-api-key): Create an Azure OpenAI resource in the Azure portal, get your API key and endpoint, deploy a model, and integrate with Bifrost for governance and cost tracking.
- [How to Get a Cerebras API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-cerebras-api-key): Generate a Cerebras API key for fast LLM inference and integrate with Bifrost for cost governance and routing.
- [How to Get a Cohere API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-cohere-api-key): Create a Cohere account, copy your trial API key from the dashboard, store it securely, and integrate with Bifrost for cost tracking, governance, and multi-provider routing.
- [How to Get a Databricks API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-databricks-api-key): Generate Databricks API credentials for foundation models and integrate with Bifrost for unified LLM governance.
- [How to Get an ElevenLabs API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-elevenlabs-api-key): Generate an ElevenLabs API key for text-to-speech and voice synthesis and integrate with Bifrost for cost tracking.
- [How to Get a Fireworks API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-fireworks-api-key): Generate a Fireworks API key for fast inference on open-source models and integrate with Bifrost for cost governance.
- [How to Get a Google Gemini API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-gemini-api-key): Create a Google account, generate your Gemini API key at Google AI Studio, set up billing for production, and integrate with Bifrost for multi-model routing and cost governance.
- [How to Get a Groq API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-groq-api-key): Create a Groq account, generate your API key from console.groq.com, set up billing, and integrate with Bifrost for cost tracking and high-speed inference routing.
- [How to Get a Hugging Face API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-hugging-face-api-key): Generate a Hugging Face access token for the Inference API and integrate with Bifrost for cost tracking and governance.
- [How to Get a Mistral API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-mistral-api-key): Create a Mistral account, generate your API key from console.mistral.ai, store it securely, and integrate with Bifrost for cost tracking and multi-provider routing.
- [How to Get Nebius API Access](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-nebius-api-key): Configure Nebius API access for open-source LLM inference and integrate with Bifrost for cost tracking.
- [How to Get an OpenAI API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-openai-api-key): Create an OpenAI API key in the dashboard, export OPENAI_API_KEY, add billing or credits for production use, and route requests through Bifrost for virtual keys, budgets, and observability.
- [How to Get an OpenRouter API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-openrouter-api-key): Generate an OpenRouter API key to route across multiple LLM providers and integrate with Bifrost for unified cost governance.
- [How to Get a Replicate API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-replicate-api-key): Create a Replicate account, copy your API token, store it securely, and integrate with Bifrost for cost tracking and governance.
- [How to Get a Runway ML API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-runway-ml-api-key): Create a Runway developer account, generate your API key from the developer portal, and integrate with Bifrost for cost tracking and governance.
- [How to Get a Vertex AI API Key](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-a-vertex-ai-api-key): Create a Google Cloud Project, enable Vertex AI API, generate a service account key, and integrate with Bifrost for cost tracking and multi-provider governance.
- [How to Set Up Ollama Locally](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-get-an-ollama-api-key): Run open-source LLMs locally with Ollama and integrate with Bifrost for local inference routing and governance.
- [How to Set Up Ollama](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-set-up-ollama): Install Ollama, pull models, and connect your local inference server to Bifrost for unified routing and observability.
- [How to Set Up vLLM](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-set-up-vllm): Deploy vLLM for high-throughput open-model serving and integrate with Bifrost for production routing and cost controls.
- [How to Set Up SGLang](https://www.getmaxim.ai/bifrost/guides/api-keys/how-to-set-up-sglang): Run SGLang for structured generation and fast serving, then connect to Bifrost for governance and multi-provider routing.
- [Claude Code guides](https://www.getmaxim.ai/bifrost/guides/claude-code): Run Claude Code through Bifrost with multi-provider routing, budgets, MCP tools, and observability.
- [Claude Code Gateway](https://www.getmaxim.ai/bifrost/guides/claude-code/claude-code-gateway): Set up Bifrost as a gateway with model routing, governance controls, and full observability.
- [Use Claude Code with OpenAI models](https://www.getmaxim.ai/bifrost/guides/claude-code/use-claude-code-with-openai-models): Route Claude Code to GPT-5 and GPT-5 mini with virtual keys, budgets, and failover.
- [Use Claude Code with Gemini models](https://www.getmaxim.ai/bifrost/guides/claude-code/use-claude-code-with-gemini-models): Route Claude Code to Gemini 2.5 Pro and Flash with virtual keys, budgets, and failover.

## Tools and data

- [Model library](https://www.getmaxim.ai/bifrost/model-library): Capabilities, context windows, and pricing for every supported model. (markdown: https://www.getmaxim.ai/bifrost/model-library.md)
- [LLM cost calculator](https://www.getmaxim.ai/bifrost/llm-cost-calculator): Estimate and compare token cost across providers. (markdown: https://www.getmaxim.ai/bifrost/llm-cost-calculator.md)
- [Provider status](https://www.getmaxim.ai/bifrost/provider-status): Live availability and incident history for supported providers. (markdown: https://www.getmaxim.ai/bifrost/provider-status.md)
- [MCP servers](https://www.getmaxim.ai/bifrost/mcp-servers): Directory of MCP servers you can connect through the MCP gateway.

## Comparisons

- [Alternatives](https://www.getmaxim.ai/bifrost/alternatives): How Bifrost compares to other AI gateways. (markdown: https://www.getmaxim.ai/bifrost/alternatives.md)
- [LiteLLM alternatives](https://www.getmaxim.ai/bifrost/alternatives/litellm-alternatives): Compare LiteLLM alternatives for scalable enterprise AI gateways, including performance, deployment complexity, governance, observability, MCP support, and migration path. (markdown: https://www.getmaxim.ai/bifrost/alternatives/litellm-alternatives.md)
- [Portkey alternatives](https://www.getmaxim.ai/bifrost/alternatives/portkey-alternatives): Compare Portkey alternatives for production AI gateways, including latency overhead, SDK compatibility, pricing complexity, MCP capabilities, deployment model, and enterprise controls. (markdown: https://www.getmaxim.ai/bifrost/alternatives/portkey-alternatives.md)

## Company

- [About us](https://www.getmaxim.ai/about-us): The team building Bifrost at Maxim.
- [Blog](https://www.getmaxim.ai/bifrost/blog): Product updates, engineering notes, and releases.
- [Articles](https://www.getmaxim.ai/articles): Long-form writing on AI infrastructure.
- [Releases](https://www.getmaxim.ai/bifrost/releases): Changelog for every Bifrost release.
- [Built with Bifrost](https://www.getmaxim.ai/bifrost/built-with-bifrost): Applications and teams running on Bifrost.
- [Partners program](https://www.getmaxim.ai/partners-program): Partner with Maxim on Bifrost deployments.
- [Careers](https://www.getmaxim.ai/careers): Open roles at Maxim.
- [Contact](https://www.getmaxim.ai/contact): Talk to the team or book a demo.
- [OSS friends](https://www.getmaxim.ai/oss-friends): Open-source projects we support.
- [Brand assets](https://www.getmaxim.ai/brand-assets): Logos and brand guidelines.

## Optional

- [Privacy policy](https://www.getmaxim.ai/privacy-policy)
- [Terms of service](https://www.getmaxim.ai/terms-of-service)