Official documentation repository for Eden AI - the unified API platform for accessing multiple AI providers through a single interface.

This repository contains the complete documentation for Eden AI's V3 API, built using Mintlify. The documentation includes guides, tutorials, API references, and integration examples for developers using Eden AI's services.

Eden AI is a unified API platform that provides:

• Universal AI Endpoint - Single endpoint for all AI features (text analysis, OCR, image processing, translation)
• OpenAI-Compatible LLM - Drop-in replacement for OpenAI's API with multi-provider support
• Smart Routing - Intelligent provider selection based on cost, performance, and reliability
• Persistent File Storage - Upload files once, use them across multiple requests
• Built-in API Discovery - Explore features and providers programmatically

• Get Started - Introduction, smart routing, FAQ, enterprise offerings, and professional services
• How-To Guides - Step-by-step guides for authentication, cost management, discovery, user management, smart routing, Universal AI, LLM endpoints, and file uploads
• Tutorials - Practical tutorials for optimizing LLM costs, tracking spending, and managing tokens
• Integrations - SDKs (Python, TypeScript), AI assistants (Claude Code, Continue.dev), frameworks (LangChain), and chat platforms (LibreChat, Open WebUI)
• Changelog - Version history and updates

OpenAPI specifications for:

• AI Features (Universal AI endpoint)
• Cost Management
• User Management

Install the Mintlify CLI:

npm i -g mint

Navigate to the documentation root directory and run:

mint dev

View the documentation at http://localhost:3000

• If the dev environment isn't running: Run mint update to get the latest CLI version
• If pages load as 404: Ensure you're in a directory with a valid docs.json file

• /v3/ - V3 API documentation pages
• /v3/features/ - Auto-generated AI feature reference pages (do not edit manually)
• /scripts/ - Automation scripts (see below)
• /tests/ - Automated test suite for Python code snippets (see tests/README.md)
• /shared/ - Reusable content snippets
• /openapi/ - OpenAPI specification files
• /images/ - Documentation images and assets
• /logo/ - Eden AI logo files (light/dark mode)
• docs.json - Mintlify configuration file
• index.mdx - Documentation homepage

The v3/features/ directory contains automatically generated documentation for all Universal AI features. These pages are created by scripts/generate_features.py, which fetches live data from the production /v3/info API.

For each AI subfeature (e.g. text/moderation, image/generation, ocr/ocr):

• An individual .mdx page with:
  • Input/output schema tables (field name, type, required, description)
  • Available providers and models with pricing
  • Quick-start code examples (Python + cURL) — built automatically from the input schema. The script picks the first available model, generates realistic placeholder values for each required field (e.g. "en" for language, "The quick brown fox..." for text), and renders ready-to-copy snippets inside Mintlify <CodeGroup> tabs. Async features get a note about polling the job endpoint.
• An index.mdx overview page with cards grouped by feature category
• The AI Features navigation group in docs.json is updated automatically

When a feature is added or removed from the API, the generator picks it up: new pages are created and stale pages are deleted.

python scripts/generate_features.py

No dependencies beyond Python stdlib. The script hits the production API at https://api.edenai.run/v3/info.

1. Run the script and verify the output:

   python3 scripts/generate_features.py

2. Check that pages were generated under v3/features/:

   ls v3/features/*/

3. Preview locally with Mintlify to verify rendering:

   npx mint dev

4. Confirm the AI Features group appears in the sidebar under V3 Documentation.

A GitHub Actions workflow (.github/workflows/generate-features.yml) runs this script daily at 06:00 UTC and on manual dispatch. If any pages changed, it opens a PR on the auto/update-feature-docs branch for review.

Input schema fields include descriptions (e.g. "Text to moderate", "ISO 639-1 language code"), but output schema fields currently have no descriptions. This is because the upstream Pydantic output models in edenai-apis don't define Field(description=...) on their fields. The generator and /v3/info API already support descriptions when present — the fix is to add Field(description="...") to the output dataclasses in edenai-apis.

The documentation is configured via docs.json, which includes:

• Navigation structure and tabs
• Theme and branding (colors, logos, favicon)
• Navbar and footer links
• OpenAPI integration
• Contextual features (copy, view, AI assistant integrations)

Changes are automatically deployed to production when pushing to the main branch. The GitHub app integration propagates changes from this repository to the live documentation site.

• Live Documentation: https://docs.edenai.co
• Eden AI Dashboard: https://app.edenai.run
• Website: https://edenai.co
• Status Page: https://app-edenai.instatus.com
• GitHub: https://github.com/edenai
• Discord: https://discord.gg/VYwTbMQc8u

For issues or questions about the documentation:

• Visit the Eden AI website
• Join the Discord community
• Check the GitHub repositories

This is the official documentation repository for Eden AI. For contributions or corrections, please contact the Eden AI team through the official channels.

Built with Mintlify