Gemini-OpenAI-Proxy

Bring OpenAI style API with low cost.

• Use Gemini models to complete OpenAI API chat and embedding calls.
• Use openai.fm to implement tts-1 model.

Introduction

Gemini-OpenAI-Proxy is a proxy software. It is designed to convert OpenAI API protocol calls into Google Gemini Pro protocol, so that software using OpenAI protocol can use Gemini Pro model without perception.

If you're interested in using Google Gemini but don't want to modify your software, Gemini-OpenAI-Proxy is a great option. It allows you to easily integrate the powerful features of Google Gemini without having to do any complex development work.

Demo

Get api key from https://makersuite.google.com/app/apikey

✅ Gemini Pro

curl -s http://localhost:8000/v1/chat/completions \
  -H "Authorization: Bearer $YOUR_GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
      "model": "gpt-3.5-turbo",
      "messages": [{"role": "user", "content": "Hello, Who are you?"}],
      "temperature": 0.7
      }'

✅ Gemini Pro Vision

curl -s http://localhost:8000/v1/chat/completions \
  -H "Authorization: Bearer $YOUR_GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "gpt-4-vision-preview",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "What do you see in this picture?"
        },
        {
          "type": "image_url",
          "image_url": {
            "url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAADAAAAAnAgMAAAA0vyM3AAAACVBMVEX/4WwCAgF3aTMcpbzGAAAAa0lEQVR4nGOgAWB1QOYEIHFEcXKmhCBxQqYgcSLEEGymAFEEhzFAFYmTwNoA53A6IDmB1YETidPAiLBVFGgEgrNqJYIzNTQU4Z5QZA6QNQ3hGpAZcNegceBOADFQOQlQDhfQyUwLkPxKVwAABbkRCcDA66QAAAAASUVORK5CYII="
          }
        }
      ]
    }
  ],
  "stream": false
}'

Plan

• /v1/chat/completions
  • stream
  • complete

Model Mappings

Request Model	Target Gemini Model
gpt-3.5-turbo	gemini-1.5-flash-8b-latest
gpt-4	gemini-1.5-pro-latest
gpt-4o	gemini-1.5-flash-latest
gpt-4o-mini	gemini-1.5-flash-8b-latest
gpt-4-vision-preview	gemini-1.5-flash-latest
gpt-4-turbo	gemini-1.5-pro-latest
gpt-4-turbo-preview	gemini-2.0-flash-exp
gemini*	gemini*
...(others)	gemini-1.5-flash-latest

Run On Serverless

Cloudflare Workers

Copy main_cloudflare-workers.mjs to cloudflare-workers

Deno Deploy

Copy main_deno.mjs to deno deploy

Vercel

• Alternatively can be deployed with cli: vercel deploy
• Serve locally: vercel dev
• Vercel Functions limitations (with Edge runtime)

Run On Local

deno

deno task start:deno

node

node dist/main_node.mjs

bun

bun dist/main_bun.mjs

docker

docker run -d -p 8000:8000 ghcr.io/zuisong/gemini-openai-proxy:deno
## or
docker run -d -p 8000:8000 ghcr.io/zuisong/gemini-openai-proxy:bun
## or
docker run -d -p 8000:8000 ghcr.io/zuisong/gemini-openai-proxy:node

Star History