Model catalog
Use the catalog to decide which model ID to send in inference requests. The public model browser and catalog endpoint show the same customer-facing data.
Endpoints
GET/v1/modelsOpenAI-compatible model list, with Anthropic family tiers for Claude clients.
GET/v1/models/{id}One model from the list.
GET/v1/models/catalogFull catalog with providers, tags, pricing, context windows, and release dates.
GET/v1/analytics/latencyPublic latency aggregates by model and provider.
Providers
Mistral AI (France)
Mistral chat, OCR, moderation, embedding, transcription and speech models, plus selected open models.
Scaleway (France)
Generative APIs for selected open models, embeddings and transcription.
OVHcloud (France)
AI Endpoints for selected open models, embeddings, moderation, transcription and image generation.
Inceptron (Sweden)
Selected large open models (GLM, Kimi, DeepSeek).
TensorX (Ireland)
Large open models (DeepSeek, GLM, Kimi, MiniMax, Qwen), embeddings, transcription and speech.
Browse models visually at /models.