Give your team a workspace. Give your apps a gateway.

One platform for your team and your code. Unified by routing, governance, and deterministic spend control.

The world's best models, unified

Why choose just one? Get instant access to leading providers like Mistral, OpenAI, and Anthropic with one account, one balance, and one set of rules.

One balance across providers, Workspace, and Gateway.

Decide what your organisation can use - by provider, model, and surface.

Give teammates spend limits and keep usage inside predictable guardrails.

Reserved-spend semantics plus caps that actually stop requests.

A premium workspace for drafting and collaboration, plus a production gateway for your code - unified by the same policies and spend controls.

Premium AI chat for your team. Private for them, governed by you.

A secure space for individual high-stakes work. Chats are strictly private to the user, no shared history, just pure focus.
Members use the specific providers and models the owner has allowlisted, without ever needing to manage keys or billing.
The owner sets the budget and policies; the workspace enforces them automatically for every member.

The production API for your code. Universal access, strict boundaries.

Full support for streaming chat and embeddings using the OpenAI SDKs your team already knows.
Your apps call explicit model IDs you choose (for example, gpt-5-mini or claude-sonnet-4-5). No hidden routing - what you request is what runs.
Budget is reserved before execution. The API enforces hard caps at the network edge, preventing runaway costs.

One platform for your team and your code. Unified by routing, governance, and deterministic spend control.

Loading…

Workspace your team loves

Premium AI chat for your team. Private for them, governed by you.

A secure space for individual high-stakes work. Chats are strictly private to the user, no shared history, just pure focus.

Members use the specific providers and models the owner has allowlisted, without ever needing to manage keys or billing.

The owner sets the budget and policies; the workspace enforces them automatically for every member.

Launch Workspace →

Gateway to 100+ AI models

The production API for your code. Universal access, strict boundaries.

Full support for streaming chat and embeddings using the OpenAI SDKs your team already knows.

Your apps call explicit model IDs you choose (for example, gpt-5-mini or claude-sonnet-4-5). No hidden routing - what you request is what runs.

Budget is reserved before execution. The API enforces hard caps at the network edge, preventing runaway costs.

View API Quickstart →