new-api vs freellmapi
Two of the top projects in Gateways, side by side: who leads on each score part and who each one suits, then setup, license, activity and what each review found.
new-api
Gateway that unifies LLM providers behind OpenAI, Anthropic and Gemini APIs
freellmapi
OpenAI-compatible router that pools free LLM provider tiers behind one endpoint
Verdict
new-api ranks higher, 82 to 77. Of the 5 score parts we measure for both, new-api leads on 2, freellmapi on 1 and 2 are even.
- Adoptionnew-api leads by 10 points, 84 to 74.
- FreshnessEven at 100.
- Maintenancefreellmapi leads by 12 points, 98 to 86.
- Easy to runEven at 67.
- Agent-readynew-api leads by 70 points, 70 to 0: a wide gap.
Pick new-api if you are
Teams and operators sharing and metering access to multiple LLM providers
- Converts between OpenAI, Anthropic Messages and Gemini request formats via RelayKit
- Channel priorities, weights, retries and multiple upstream keys per channel
Pick freellmapi if you are
Developers and self-hosters stretching free LLM tiers across agents
- Per-key rate counters and automatic failover on 429/5xx across providers
- Covers chat, embeddings, images, video, speech, Anthropic, Gemini and Ollama APIs
| What we compare | new-api | freellmapi |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 84, widely used | 74, popular |
| Freshness | 100, active | 100, active |
| Maintenance | 86, healthy | 98, healthy |
| Easy to run | 67, easy | 67, easy |
| Agent-ready | 70, partly | 0, none |
| Facts from GitHub and the README | ||
| Stars | 50.1k | 36.2k |
| License | AGPL-3.0 (copyleft) | MIT (permissive) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | Oct 2026 | Oct 2026 |
| Language | Go | TypeScript |
| Docker | Yes | Yes |
| GPU | Not needed | Not needed |
| arm64 or Apple Silicon | Mentioned | Mentioned |
new-api
New API is a Go and Gin gateway that connects upstream providers such as OpenAI, Anthropic, Gemini, Azure OpenAI, AWS Bedrock, Vertex AI, DeepSeek and Qwen. It exposes them through OpenAI Chat/Responses, Anthropic Messages and Gemini endpoints. A web console handles channel routing, users, quotas, pricing and usage logs. It runs on SQLite by default, or on MySQL or PostgreSQL with optional Redis.
Who it is for: Teams and operators sharing and metering access to multiple LLM providers
Strengths
- Converts between OpenAI, Anthropic Messages and Gemini request formats via RelayKit
- Channel priorities, weights, retries and multiple upstream keys per channel
- Per-user quotas, subscriptions, usage logs and expression-based pricing
- Starts on SQLite with one container; Compose file adds PostgreSQL and Redis
Weaknesses
- AGPL-3.0 with extra terms requiring attribution and a visible project link in modified UIs
- Protocol conversion is not exact; tools and fields may not map across protocols
- Multi-node setups need shared database, secrets and Redis, or rate limits count per node
- Public or resale use brings filing and compliance duties the README leaves to the operator
- no GPU
- Docker + Compose
- Needs SQLite, MySQL >= 5.7.8 or PostgreSQL >= 9.6, Redis (optional), ClickHouse (optional, log database)
- Models: OpenAI, Anthropic, Google Gemini, Azure OpenAI, AWS Bedrock
- port 3000
freellmapi
FreeLLMAPI exposes a single /v1 API (chat, responses, embeddings, images, audio, plus Anthropic Messages, Gemini and Ollama surfaces) and routes requests across free tiers from 34 providers. It stores provider keys AES-256-GCM encrypted in SQLite, tracks RPM/RPD/TPM/TPD per key, and fails over to the next model on 429/5xx. A React dashboard manages keys and the fallback chain; a desktop app is also available.
Who it is for: Developers and self-hosters stretching free LLM tiers across agents
Strengths
- Per-key rate counters and automatic failover on 429/5xx across providers
- Covers chat, embeddings, images, video, speech, Anthropic, Gemini and Ollama APIs
- Keys encrypted at rest in SQLite; clients only see one unified token
- Setup commands for Claude Code, Codex, Cursor, Aider and other agents
Weaknesses
- Free installs get new models 30 days late; same-day catalog needs $19/yr premium
- Local-first and single-user by design; no multi-user support
- Depends on free-tier quotas that providers change without notice
- Model catalog sync pulls a signed feed from freellmapi.co
- no GPU
- Docker + Compose
- Needs Node.js 20+, SQLite, Docker (optional), provider API keys
- Models: OpenAI-compatible APIs, Anthropic Messages, Gemini API, Ollama API, custom OpenAI-compatible endpoints
- port 3001
More in Gateways
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.