AIWave
Create API key

One endpoint. Every model behind it.

AIWave sits between your client and eight model families. One key, one header, one routing layer. Your requests come in; the work comes back.

Create API key See the families
Your client Unchanged One key One header AIWave Route · auth · rate basis GPT Claude Gemini DeepSeek GLM Kimi Qwen Visual Swappable
How a request travels — the client stays as it is; the routing happens here.

08 families

GPTClaudeGeminiDeepSeekGLMKimiQwenVisual

Eight families, one key.

Every family answers through the same surface. What each is for, in one line each — the full list lives on Models.

  • GPTText & image · OpenAI-compatible
  • ClaudeReasoning · two contracts
  • GeminiMultimodal input · native format
  • DeepSeekReasoning & code · OpenAI-compatible
  • GLMProduction work · OpenAI-compatible
  • KimiLong context · OpenAI-compatible
  • QwenText & multimodal · OpenAI-compatible
  • VisualStill image & motion · generation paths

Read the full model landscape →

Two contracts, not ten.

Almost every family answers to the OpenAI-compatible shape. The exceptions are deliberate: Gemini's native input format, Claude's native Messages contract, and the asynchronous generation paths.

Compatible
/v1/chat/completions
Image
/v1/images/generations
Video
/v1/video/generations
Auth
Bearer key, one header

Swap the model. Keep the client.

No rates, uptime figures, or model identifiers are quoted on this page — they move, so they live at the point of use.

Start in five minutes See what it costs