GPT-4o, o3, o4-mini, DALL-E, Whisper. Drop-in proxy with PII shield.
Claude Opus, Sonnet, Haiku. Full Messages API with auto-tokenisation.
Gemini 2.5 Pro/Flash, 2.0 Flash. generateContent proxy with PII interception.
Grok-3, Grok-3-mini, Grok-2-vision. OpenAI-compatible SDK.
200+ models via single endpoint — Llama, Mistral, Qwen, Command R, and more.
Mistral Large, Small, Codestral. European sovereign AI option.
DeepSeek-V3, DeepSeek-R1. High-performance reasoning models.
LPU inference — ultra-low latency for Llama and Mixtral.
Local inference for any GGUF model. On-premises sovereign deployment.
Inference Endpoints and Hub model access. Open-weight model library.
Search-augmented AI with live citations. Web-grounded responses.