HopBase DocsAPI overview
Call chat, image, video, and speech models through one Base URL; examples are tested daily.
Developer quickstart
Send your first request in minutes. Keep using the OpenAI or Anthropic SDK — just change the Base URL and API key.
curl https://api.hop-base.com/v1/chat/completions \
-H "Authorization: Bearer $HOPBASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"messages": [
{
"role": "user",
"content": "Write a one-line product tagline"
}
]
}'Start here
Choose how to call HopBase and a model family
Will you use a coding tool such as Claude Code or Codex, or call the API from your own code? Decide that first, then pick a model family.
Create an API key in the right group
Each key belongs to exactly one group, and the group decides which models the key can call. To use several model families, create one key per family.
Install an SDK, or send plain HTTP
The official SDKs work as is — change only
base_urland the key. No SDK? Use curl.Send your first request
Follow the quickstart to complete one chat request and confirm that the key, group, and Base URL match.
Before you go to production
Learn the error codes and retry rules, plus concurrency and timeout limits. When troubleshooting, include the
x-request-idresponse header.Error codes and retriesConcurrency, timeouts, and billingTroubleshooting
Start with a task
- Generate product imagesText to image, edits with reference images, several images per request.GPT Image · Gemini · Seedream
- Set up Claude Code or CodexThe console generates one command. Three steps in your terminal.macOS · Linux · Windows
- Call the API from your appKeep the OpenAI or Anthropic SDK. Change only the Base URL and API key.Python · Node.js · Go · Java
- Generate videoText to video, first and last frames, reference media. Submit, then poll for the result.Seedance · Kling · MiniMax · Wan
- Text to speech and voice cloningUse the official voices as they are, or clone your own.MiniMax Speech 2.8
- Migrate from OpenAI or OpenRouterChange the Base URL and API key, and the model ID if needed.One API key per group
Not sure which image model to use? Answer three questions
- 1What are you making?
- 2What matters most?
- 3Largest size
gemini-3.1-flash-imageSupports 1K / 2KGemini 3.1 Flash Image
Flash tier at $0.0672 / image, up to 14 reference images, and moving up to 2K costs nothing extra
Recommendations are based on official prices and specs, not on image-quality benchmarks.
Other entry points
| What you want to do | Start here |
|---|---|
| Have a coding AI integrate HopBase into an existing project | Integrate with AI |
| Create images and videos in the browser, no code | Creative Studio |
| Manage Claude Code and Codex settings in a desktop app | CC Switch |
| Configure Cursor, Cline, Roo Code, or Continue | Editors and IDEs |
| Configure Cherry Studio, NextChat, or LobeChat | Third-party clients |
| Compare official prices, limits, and availability across models | Models |
| Generate image requests and code from each model's spec | Image request builder |
| Fix 401, 404, model-not-found, and other errors | Troubleshooting |