# WaveSpeedAI

> WaveSpeedAI is a unified AI inference platform offering 1,000+ image, video, audio, and 3D generation models alongside 290+ LLMs through a single OpenAI-compatible API. Pay-per-token pricing, no cold starts, transparent per-model rates.

## Core API

- [All Models](https://wavespeed.ai/models): Browse 1,000+ image, video, audio, and 3D models — Seedance 2.0, Kling 3.0, Veo 3.1, Nano Banana 2, Seedream, Flux, Sora 2, and more.
- [API Documentation](https://wavespeed.ai/docs): Quick-start guides for Python, JavaScript, and cURL.
- [CLI](https://wavespeed.ai/cli): Open-source command-line client (`npm install -g @wavespeed/cli`). Run any model from your terminal or wire it into an AI coding agent (Codex, Claude Code, Claude Cowork, Cursor). Source: github.com/WaveSpeedAI/wavespeed-cli.
- [Pricing](https://wavespeed.ai/pricing): Per-token rates for every model. No subscriptions; only pay for what you use.

## AI Media Generators

- [Image Generator](https://wavespeed.ai/image-generator): Multi-model image generation in the browser — Nano Banana 2/Pro, Seedream 4.5, GPT Image 2, Qwen Image 2.0, WAN 2.7, FLUX 2 Klein.
- [Video Generator](https://wavespeed.ai/video-generator): Text-to-video and image-to-video in the browser — Seedance 2.0, WAN 2.7, Kling 3.0/O3, Veo 3.1, Hailuo 2.3, Vidu Q3, Sora 2.
- [Avatar Generator](https://wavespeed.ai/avatar-generator): Talking-head and avatar synthesis.
- [Audio Generator](https://wavespeed.ai/audio-generator): TTS and music generation.
- [3D Generator](https://wavespeed.ai/3d-generator): Image-to-3D and text-to-3D mesh generation.

## Image Models

- [Nano Banana 2](https://wavespeed.ai/models/google/nano-banana-2/text-to-image): Google's flagship image model — high prompt fidelity for text-to-image and editing.
- [Nano Banana Pro](https://wavespeed.ai/models/google/nano-banana-pro/text-to-image): Pro tier of Google Nano Banana with higher resolution and detail.
- [Seedream 4.5](https://wavespeed.ai/models/bytedance/seedream-v4.5): ByteDance Seedream — strong photoreal and stylized image generation.
- [WAN 2.7](https://wavespeed.ai/models/alibaba/wan-2.7/text-to-image): Alibaba WAN 2.7 image — text-to-image and edit, plus a Pro variant.
- [GPT Image 2](https://wavespeed.ai/models/openai/gpt-image-2/text-to-image): OpenAI GPT Image 2 for text-to-image and edit.
- [Qwen Image 2.0](https://wavespeed.ai/models/wavespeed-ai/qwen-image-2.0/text-to-image): Alibaba Qwen Image 2.0, with a Pro variant.
- [FLUX 2 Klein](https://wavespeed.ai/models/wavespeed-ai/flux-2-klein-9b/text-to-image): Black Forest Labs FLUX 2 Klein 9B — open-weights image model.
- [Z-Image Turbo](https://wavespeed.ai/models/wavespeed-ai/z-image/turbo): Fast text-to-image model.

## Video Models

- [Seedance 2.5 Coming Soon](https://wavespeed.ai/seedance-2-5-api): ByteDance Seedance 2.5 preview page. Coming in Early July with planned 30-second single-shot output, larger reference capacity, more controllable generation/editing, and up to 50 reference files; live examples currently use Seedance 2.0 endpoints.
- [Seedance 2.0](https://wavespeed.ai/models/bytedance/seedance-2.0/image-to-video): ByteDance Seedance 2.0 — text-to-video, image-to-video, and video-edit; up to 1080p.
- [Seedance 2.0 Turbo](https://wavespeed.ai/models/bytedance/seedance-2.0/image-to-video-turbo): Faster Seedance 2.0 variant for quick iteration.
- [WAN 2.7](https://wavespeed.ai/models/alibaba/wan-2.7/image-to-video): Alibaba WAN 2.7 video — text-to-video, image-to-video, and video-edit.
- [Kling 3.0](https://wavespeed.ai/models/kwaivgi/kling-v3.0-pro/image-to-video): Kuaishou Kling 3.0 (std + pro) — text-to-video and image-to-video.
- [Kling O3](https://wavespeed.ai/models/kwaivgi/kling-video-o3-pro/image-to-video): Kling O3 (std + pro) — text/image/reference-to-video and video-edit.
- [Veo 3.1](https://wavespeed.ai/models/google/veo3.1/image-to-video): Google Veo 3.1, plus Fast and Lite variants.
- [Hailuo 2.3](https://wavespeed.ai/models/minimax/hailuo-2.3/i2v-standard): MiniMax Hailuo 2.3 — t2v and i2v (standard + pro).
- [Vidu Q3](https://wavespeed.ai/models/vidu/q3/image-to-video): Vidu Q3 video generation.
- [Grok Imagine 1.5](https://wavespeed.ai/models/x-ai/grok-imagine-video-v1.5/image-to-video): xAI Grok Imagine 1.5 image-to-video.
- [Grok Imagine](https://wavespeed.ai/models/x-ai/grok-imagine-video/text-to-video): xAI Grok Imagine.
- [Sora 2](https://wavespeed.ai/models/openai/sora-2/text-to-video): OpenAI Sora 2 — text-to-video and image-to-video, with a Pro variant.

## Audio Models

- [Qwen3 TTS](https://wavespeed.ai/models/wavespeed-ai/qwen3-tts/text-to-speech): Alibaba Qwen3 TTS — text-to-speech, voice clone, and voice design.
- [OmniVoice](https://wavespeed.ai/models/wavespeed-ai/omnivoice/text-to-speech): WaveSpeedAI OmniVoice — TTS and voice cloning.
- [ElevenLabs v3](https://wavespeed.ai/models/elevenlabs/eleven-v3): ElevenLabs v3 expressive TTS.
- [MiniMax Speech 2.6](https://wavespeed.ai/models/minimax/speech-2.6-hd): MiniMax Speech 2.6 (turbo + HD).
- [ElevenLabs Music](https://wavespeed.ai/models/elevenlabs/music): ElevenLabs music generation.
- [MiniMax Music Cover](https://wavespeed.ai/models/minimax/music-cover): MiniMax music cover generator.

## 3D Models

- [Tripo3D 2.5](https://wavespeed.ai/models/tripo3d/v2.5/image-to-3d): Tripo3D 2.5 — image-to-3D and multiview-to-3D.
- [Hunyuan3D 3.1](https://wavespeed.ai/models/wavespeed-ai/hunyuan-3d-v3.1/image-to-3d-rapid): Tencent Hunyuan3D 3.1 — text/image-to-3D rapid.
- [Hunyuan3D 3](https://wavespeed.ai/models/wavespeed-ai/hunyuan3d-v3/image-to-3d): Hunyuan3D 3 — text, image, and sketch-to-3D.
- [Meshy 6](https://wavespeed.ai/models/wavespeed-ai/meshy6/image-to-3d): Meshy 6 with quad/triangle topology and PBR.
- [Hyper3D Rodin 2](https://wavespeed.ai/models/hyper3d/rodin-v2/image-to-3d): Hyper3D Rodin 2 — high-quality mesh export (GLB/FBX/OBJ/USDZ).

## Avatar / Motion Models

- [InfiniteTalk](https://wavespeed.ai/models/wavespeed-ai/infinitetalk): Talking-head and video-to-video lip-sync.
- [WAN 2.2 Animate](https://wavespeed.ai/models/wavespeed-ai/wan-2.2/animate): Animate a static image into video.
- [Kling 3.0 Motion Control](https://wavespeed.ai/models/kwaivgi/kling-v3.0-pro/motion-control): Reference-driven motion control.
- [Face Swapper](https://wavespeed.ai/models/wavespeed-ai/image-face-swap): Image and video face swap.

## LLM API

- [LLM API](https://wavespeed.ai/llm): 290+ language models — GPT-5.5, GPT-5.4, Claude Opus 4.7, Gemini 3, Qwen 3, DeepSeek R1, Llama 4, Grok 4, Mistral, and more. OpenAI-compatible endpoint at `https://llm.wavespeed.ai/v1`.
- [LLM Playground](https://wavespeed.ai/llm/playground): Live test and compare LLMs in the browser — see responses, latency, and cost in real time.

## Tools and Apps

- [Tools](https://wavespeed.ai/tools): Single-purpose utilities — upscaler, watermark remover, background remover, and more.
- [Apps](https://wavespeed.ai/apps): Specialized creative apps — virtual try-on, image translator, camera-angle, and more.
- [Ad Templates](https://wavespeed.ai/apps/ads): AI-generated ad creatives across 20+ formats.

## Account

- [Sign In](https://wavespeed.ai/sign-in): Account access.
- [Dashboard](https://wavespeed.ai/dashboard): Usage, billing, and request logs.
- [Access Keys](https://wavespeed.ai/accesskey): Create and manage API keys.
- [Top Up](https://wavespeed.ai/top-up): Credit purchases via Stripe, PayPal, Airwallex, WeChat Pay/Alipay, and KRW Pay.

## Company

- [About](https://wavespeed.ai/about)
- [Enterprise](https://wavespeed.ai/enterprise): Custom deployments, SLAs, and contact sales.
- [Blog](https://wavespeed.ai/blog): Model launches, integrations, and engineering posts.
- [Affiliate](https://wavespeed.ai/static/affiliate)

## Optional

- [Terms of Service](https://wavespeed.ai/static/terms)
- [Privacy Policy](https://wavespeed.ai/static/privacy)
- [Return Policy](https://wavespeed.ai/static/return-policy)
- [Support](https://wavespeed.ai/user/support)