OSS Tanbou

centralize multiple LLM providers behind one high-performance gateway

About these scores

OSS scale score is an unbounded metric that log-compresses and weights Stars, Watchers, Forks, and Contributors. Discovery score is the current OSS scale score minus the score at discovery. Update pace is commits in the last 30 days, growth momentum is the OSS scale score difference within the recent observation window, and OSS health is a 0–100 rating based on available recency, Community Health, and release data.

Stars
8,459
Primary language
Go
License
Apache-2.0
Repository last updated
Sep 30, 2026
On this page

Overview

Bifrost is a Go AI gateway that exposes an OpenAI-compatible API while centralizing provider routing, fallbacks, load balancing, semantic caching, governance, and observability.

Features and best fit

Based on official documentation; not hands-on tested · Content checked:

Key features

Route one API to multiple LLM providers

OpenAI, Anthropic, AWS Bedrock, Google Vertex, and other providers can sit behind one application-facing interface.

Sources: [2]

Use fallbacks, load balancing, and caching for reliability and cost control

Provider and model failures, request distribution, and semantic caching can be handled at the gateway layer.

Sources: [2]

Best fit

Fits teams separating LLM routing and observability from application code

Multiple applications can share routing, credentials, usage controls, and monitoring through a common gateway.

Sources: [2]

Before adoption

Gateway releases and the CLI use different version lines

The HTTP transport release is v2.2.4 while the NPX CLI package metadata reports 1.6.3. Pin the distribution you actually deploy.

Sources: [4][1]

Check the boundary of Enterprise deployment capabilities

The README advertises private networking, custom security controls, clustering, and related capabilities under Enterprise Deployments. Verify whether required features are available in the OSS distribution.

Sources: [2]

Official sources

  1. [1]maximhq/bifrost repository(2026-09-30)
  2. [2]Bifrost README(2026-09-30)
  3. [3]Bifrost Apache-2.0 license(2026-09-30)
  4. [4]Bifrost HTTP v2.2.4 release(2026-09-30)
Supplemental curator note

Bifrost is a Go AI gateway that unifies OpenAI, Anthropic, Bedrock, Vertex AI, and other providers behind a common API. The repository is Apache-2.0 and the latest HTTP transport release is v2.2.4. The README separately advertises Enterprise Deployments with additional capabilities such as private networking, custom security controls, and clustering, so verify feature availability before adoption.

Try it in 3 steps

  1. 1

    Check Node.js 18+

    Verify the runtime requirement for the NPX CLI package.

    node --version
  2. 2

    Start the Bifrost gateway

    Pin the NPX CLI and start the gateway and Web UI on localhost:8080. The HTTP transport uses a separate version line.

    npx -y @maximhq/bifrost@1.6.3
  3. 3

    Configure a provider and test the API

    Set a provider and API key in the Web UI, then try the OpenAI-compatible /v1/chat/completions endpoint.

    open http://localhost:8080
Check the official README

Growth

Growth trends · Last 30 days

8,459 Stars

Trend data is still being collected.

Built with

Categories and tags

GitHub data

GitHub dataView detailed GitHub data

GitHub Topics

  • ai-gateway
  • gateway
  • gateway-services
  • generative-ai
  • guardrails
  • llm
  • llm-cost
  • llm-gateway
  • llm-observability
  • llmops
  • load-balancing
  • mcp-client
Stars
8,459
Forks
1,313
Watchers
33
Open issues
1,150
Owner type
Organization
Primary language
Go
License
Apache-2.0
Repository last updated
Sep 30, 2026
Write a related article

Share a guide or use case for this OSS in Markdown. Articles are published after administrator approval.

Report incorrect information

Tell us if any listing information is incorrect or outdated.

After reading this page, do you know what to do next?