Bifrost Gateway

Bifrost Gateway is a high-performance AI gateway for self-hosted Kubernetes workloads that routes calls to 23+ model providers through a single…

  • AI
  • Development
  • Tools & Utilities
Bifrost Gateway is a high-performance AI gateway for self-hosted Kubernetes workloads that routes calls to 23+ model providers through a single OpenAI-compatible API, simplifying provider management and reducing downtime.

Platform availability

Available on 2 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

81/100 Excellent

Maintained
100last commit 14 days ago
Popular
799,424 GitHub stars
Easy to install
452 install platforms, install notes, Docker image
Light to run
No data

Based on 3 of 4 factors.

Availability: platform listings confirmed today

GitHub stars
9,424
Open issues
1453
Last commit
2026-11-07
Latest release
ent-v2.5.0-base
License
Apache-2.0
Activity
rising

Checked 14 days ago - source: GitHub (maximhq/bifrost)

Resources & Compatibility

ARM (e.g. Raspberry Pi) is supported.

Checked 11 days ago - source: Docker Hub image tags and published docs

First-install notes

  • If using Helm, start with the guided installer From the helm-charts/bifrost folder, run ./scripts/install.sh. It walks through the setup choices for you, which helps avoid missing a required setting and ending up with a broken install. source
  • Install from an example Helm config When running helm install, point to an example values file such as postgres-only.yaml and set the image tag to the version you want. That gives Bifrost the database connection and exact release it expects, instead of a half-filled config that won't start. source
  • Set secrets, TLS, and network rules for HA For the HA example, create the needed Kubernetes secrets (the stored database password and encryption key), apply the permission rules it expects, and configure TLS. Also let the pods talk to each other on gossip port 7946, which Bifrost uses for replica health checks; without these pieces, the app may fail to start or replicas won't connect securely. source

Fetched 41 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Can Bifrost present multiple LLM providers through a single OpenAI-compatible API?

Yes. The repository says it unifies access to 23 or more providers, including OpenAI, Anthropic, AWS Bedrock, and Google Vertex, through one OpenAI-compatible API.

Does Bifrost include automatic failover and load balancing?

Yes. The project description lists automatic failover and load balancing among the gateway features you get with deployment.

Can the Bifrost AI Gateway be self-hosted with Docker in air-gapped environments?

Yes. The Docker self-hosting guide covers deployment from a quick-start Compose setup through air-gapped enterprise deployments.

Does the Docker self-hosting guide mention persistent storage?

Yes. It states that the self-hosted Docker deployment provides full persistence.

Does Bifrost have MCP support for agent workflows?

The LLM gateways comparison highlights MCP support as one of Bifrost's claimed capabilities for production or agentic LLM routing.

Answers sourced from codexpedite.com, github.com, www.reddit.com

Community

Collected 48 days ago - public community threads