Your AI providers and your enterprise stack — governed.

Claude's full model lineup — including Opus 4.6, Sonnet 4.6, and Haiku 4.5. Bring your own Anthropic API key and OBEL routes every request through the full governance pipeline.

GPT-5.5, GPT-5.4, GPT-4.1, o3, and more. Connect your OpenAI account and OBEL routes requests while enforcing your organisation's governance rules — PII scrubbing and classification run before every call.

Gemini 2.5 Pro, 2.5 Flash, and the full Gemini family. Connect your Google AI Studio key and access Google's frontier models through OBEL's governed interface.

GPT-4o, o3, and o4-mini served from your own Azure subscription. Use your Microsoft Enterprise Agreement credits, keep data inside your Azure tenant, and connect to OBEL with no new vendor procurement.

Meta Llama models served at high speed via Groq's LPU inference engine. No API key required — OBEL provides access. All models still run through the full governance pipeline.

OBEL-managed model bundles — no API key required. OBEL Free routes to fast Groq-powered Llama models at no cost. OBEL Paid gives your team access to frontier models (GPT-5.4, Claude Sonnet 4.6, Gemini 2.5 Pro) under a single prepay subscription.

Bring your own model endpoints — Ollama, vLLM, LocalAI, or any OpenAI-compatible API. Register a custom endpoint in Settings and OBEL routes your team's requests with full governance applied.

AI purpose-built for legal research, contract analysis, and due diligence. Governed access to Harvey's specialised legal models through OBEL's classification and audit pipeline.

Cohere's Command and Embed models — purpose-built for enterprise RAG, semantic search, and retrieval pipelines. Bring your Cohere key and OBEL governs every embedding and generation call.

France-based frontier models with GDPR-native data handling and EU-resident infrastructure. Connect Mistral Large and Mixtral for teams with European sovereignty requirements.

Real-time web search and citation-grounded answers. Govern Perplexity's research models through OBEL — scrubbing applied to queries before they reach the live web index.

xAI's Grok 3 family — large context, real-time data access. Bring your xAI API key and OBEL applies the same governance pipeline as every other provider: scrub, classify, audit, infer.

Using a provider not listed here? Let us know. We prioritise integrations based on demand — legal, medical, defence, and sector-specific models are especially welcome.

One pipeline, every provider

Switching providers does not change your governance posture. PII scrubbing, sovereign classification, cost governance, and the immutable audit write run on every request — regardless of which model or provider handles inference. Your compliance obligations are met at the gateway, not at the provider.