Eden AI (France) is an AI gateway that gathers models from dozens of providers into a single API. Beyond language models, it covers modalities such as image, audio, video, OCR and translation. The gateway standardizes requests, responses, authentication, billing and monitoring; inference still runs at the origin providers.
What it’s for
- Accessing hundreds of models from many providers through one endpoint
- Comparing and switching providers without rewriting the integration
- Consolidating usage split across providers into a single invoice
- Routing specialized tasks such as OCR, text analysis, image and audio
- Monitoring cost, latency and usage per key
How it works
- The application calls Eden AI’s v3 API, which standardizes input and output formats
- The gateway picks the provider according to configuration and applies fallback when needed
- Your own provider keys can replace platform credits
- The European endpoint keeps processing inside the European Union
- Eden AI acts as a technical intermediary: it does not operate third-party models
How it differs
- Multimodal breadth is the core of the product, not just text
- Unified billing covers different providers in one account
- It is an aggregation and forwarding layer, not a serving platform
- Data retention is zero by default, according to the official documentation
Availability and pricing
- Pay-as-you-go at the provider’s list price, plus a platform fee
- Enterprise with service-level agreements, dedicated keys and custom deployments
Things to note
- Each model’s availability and behavior depend on the origin provider
- Migrating between API versions requires following the changelog
- Part of the catalog is specialized and has uneven language coverage