Generate your first image
Follow the path from account setup to your first image generation request.
Build with HiAPI
One API key for Nano Banana, Flux, GPT Image, Kling, Seedance and more — OpenAI-compatible, ready in minutes. Connect AI tools through MCP, or install Skills when an agent should call one model directly.
Start Using the API
Section titled “Start Using the API”API keys and auth
Get an API key, learn request authentication, and keep production keys safe.
SDKs and examples
Use Python, Node.js, or cURL with OpenAI-compatible request formats.
Skills
Section titled “Skills”Skills are local single-model packages for agent runtimes. They are useful when an agent should call one model directly and save the generated result. GPT Image 2 and HappyHorse 1.0 Video Skills are available for Codex, Claude Code, OpenCode, OpenClaw, and similar agents.
GPT Image 2 Skill
Let an agent call HiAPI gpt-image-2 directly for a fixed text-to-image workflow.
HappyHorse 1.0 Video Skill
Let an agent call HiAPI happyhorse-1-0 directly for a fixed text-to-video workflow.
Plug into AI Tools (MCP)
Section titled “Plug into AI Tools (MCP)”Connect HiAPI to Claude Desktop, Claude Code, Codex, Cursor, VS Code/GitHub Copilot, Cherry Studio, and other MCP hosts. Add the HiAPI remote URL with your API key, then ask your AI tool to generate images, edit images, create videos, inspect models, or check pricing directly from chat.
Claude Desktop
Add the HiAPI URL and API key in Connectors, then generate images and videos from chat.
Claude Code (CLI)
Connect HiAPI in Claude Code so your terminal AI assistant can call image and video models.
Codex
Add HiAPI to Codex CLI or Desktop and generate assets while you work.
Cursor
Add HiAPI in Cursor Settings → MCP and generate assets while you code.
VS Code / GitHub Copilot
Connect HiAPI in Copilot Agent Mode so Copilot can call image and video tools.
Cherry Studio
Add HiAPI in the desktop AI client and generate images or videos through chat.
Image Generation Models
Section titled “Image Generation Models”Nano Banana series
By Google. Ultra-fast via chat/completions. Best for conversational iteration and batch generation.
FLUX 1.1 Pro
By Black Forest Labs. Photorealistic output — perfect for ads and product photography.
GPT Image 1.5 / 2
By OpenAI. Best for text rendering, complex prompts, and layout-heavy posters.
Qwen Image 2.0
By Alibaba. Best Chinese text rendering, 2K output, cost-effective.
Video Generation Models
Section titled “Video Generation Models”Seedance series
By ByteDance. 1080P narrative video with native audio. Pro and 1.0 Standard available.
Kling v3
By Kuaishou. Cinematic short videos with strong motion dynamics.
Wan 2.7 T2V / I2V
By Alibaba. Supports both text-to-video and image-to-video workflows.
API Reference & Guides
Section titled “API Reference & Guides”Image Generation Guide
Both API formats (Images API and Chat Completions), sizes, and prompt tips.
Video Generation Guide
Async polling flow, image-to-video, and complete code examples.
API Endpoints
Image, video, and account endpoints — plus error codes and rate limits.
Resources for AI Tools
Section titled “Resources for AI Tools”If you are building an AI Agent or want LLMs to quickly understand HiAPI, provide these machine-readable resources:
llms.txt
Plaintext API and tool context optimized for LLM context windows.
OpenAPI Spec
Machine-readable OpenAPI 3.0 spec for generated SDKs, Postman, and tool integrations.