Skip to main content

​
InfinityBlue API

InfinityBlue is a unified, OpenAI-compatible API gateway. Send a single request format to a single endpoint and reach every major frontier model — GPT, Gemini, DeepSeek — without managing separate provider accounts, keys, or SDKs.

​
What you get

Quickstart

Get your first request working in 5 minutes

Model Selection

Compare every supported model and pick the right one

Streaming

Return tokens to users as soon as they are produced

Multimodal Input

Send images alongside text in a single turn

​
Authentication

All requests require a Bearer token in the Authorization header: 
curl https://api.getinfinityblue.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
See Authentication for instructions on obtaining and managing your API key.

​
Supported models

The table below is a quick reference. For capability details and pricing see Model Selection and the live pricing page.

​
Chat & reasoning

ModelProviderNotes
gpt-5.4OpenAIFlagship — best overall quality
gpt-5.4-miniOpenAIFast and cost-efficient
gpt-5.5OpenAILatest generation
gpt-5.3-codexOpenAIOptimised for code tasks
gemini-3.1-pro-previewGoogleLong context, strong reasoning
gemini-2.5-proGoogleHigh-quality multimodal
gemini-2.5-flashGoogleFast and affordable
gemini-3-flash-previewGoogleLatest Flash generation
deepseek-v4-proDeepSeekAdvanced reasoning
deepseek-v4-flashDeepSeekLightweight DeepSeek

​
Image generation

ModelNotes
gpt-image-2OpenAI image generation
nanobananaFast image generation
nanobanana_proHigh-quality image generation
nanobanana_2Latest Nano Banana generation

​
Video generation

ModelNotes
doubao-seedance-2-0-260128Bytedance video generation
veo_3_1Google Veo video generation
kling-v2-5-turboKling fast video generation
If you find anything missing or unclear in these docs, open an issue or PR on GitHub.