Hyperouter is generally available

October 25, 2026 · Hyperouter · 3 min read

Hyperouter is generally available. It is one HTTP API in front of several model vendors: you send a request that names a model, Hyperouter sends it to a provider that can serve it, and your balance is charged at that provider's list price, with no markup.

What you can call today

• Chat: xAI's Grok models, from x-ai/grok-4.7 to x-ai/grok-build-0.1. DeepSeek V4 Pro and V4 Flash are coming soon.
• Images: Grok Imagine (x-ai/grok-imagine-image, with its 2.0 and quality versions) and Qwen Image 2.0 (qwen/qwen-image-2.0), to generate and edit pictures.
• Video: Grok Imagine Video (x-ai/grok-imagine-video, x-ai/grok-imagine-video-1.5 and x-ai/grok-imagine-video-1.5-lite), from a prompt, a picture or a video.

GET /v1/models returns the models your account can call at the time of the request, with prices. The docs list every model and its price.

OpenAI-compatible

Every endpoint follows OpenAI's request and response shapes. Point an OpenAI SDK at https://api.hyperouter.app/v1, use your Hyperouter key, and your code works. Hyperouter adds a few fields: models for fallback models, provider for routing preferences, and usage.cost, the charge of each request in USD. The guide to the OpenAI SDK shows the first calls.

One balance, billed at list price

• Requests are paid from a prepaid USD balance, at the list price of the provider that served them. There is no markup.
• A failed request costs nothing, and neither does a fallback attempt that fails.
• Every response carries its charge as usage.cost. The dashboard shows every request and your spend by day, model and key.
• Top-ups are manual for now: open a top-up request in the dashboard, and staff credit your balance once the payment is confirmed. Each top-up carries a fee of 5.5% of the amount paid, with a minimum of $0.80.

Accounts and keys

Sign in at dash.hyperouter.app with Google or GitHub. One account holds one balance and any number of named API keys. Each key can have its own credit limit and can be disabled at once. A key is shown once, when you create it: Hyperouter stores only its hash.

Rate limits per account

Each account has a limit on requests per minute and on requests in flight at once, shared by all its keys. The limits rise with the total you have topped up, starting at 60 requests per minute and 10 concurrent requests. The rate limits section lists the tiers, and your current limits are on the Credits page of the dashboard.

Docs for people and for SI assistants

The API reference covers every endpoint, error code and parameter. It is also published as Markdown: /llms.txt is the index for SI assistants, and /llms-full.txt holds every page in one file.

Questions? Open a ticket in the dashboard or write to [email protected].