toolingfor.com

OpenRouter

OpenRouter is a hosted API that routes requests to language models from many providers through one endpoint and one prepaid credit balance.

OPENROUTER.AI ↗ DOCUMENTATION ↗

What OpenRouter does

OpenRouter gives access to models from many providers through a single API endpoint, and the OpenAI SDK can be pointed at it. The customer buys credits in US dollars; each request is charged at the provider's list price and deducted from the balance. When a provider fails, the request can fall back to another, and an attempt that errors is not charged. Customers can also bring their own provider keys, in which case the provider bills them directly.

It is a hosted service only. By default it does not store prompts or responses; it stores request metadata such as token counts and latency, and logging of prompt content is opt-in. On the Business and Enterprise plans, requests sent to a regional endpoint are processed within the EU or the US.

OpenRouter announced on 19 August 2026 that it is joining Stripe. The announcement says it will continue to operate under the same name, with the same product, and that nothing changes for existing integrations.

OpenRouter: owner, license and hosting

Fact Value Source and date checked
Owner OpenRouter, which announced on 19 August 2026 that it is joining Stripe. SOURCE ↗
License Proprietary. A hosted service owned and operated by OpenRouter; no source license. SOURCE ↗
Self-hosting No. The documentation describes a hosted API only and no self-hosted version. SOURCE ↗
Hosted regions EU and US regional endpoints (eu.openrouter.ai, us.openrouter.ai) on the Business and Enterprise plans. The region of the default endpoint is not stated. SOURCE ↗

Each value was read from the vendor's own page on the date shown. How listings are checked

OpenRouter plans and prices

OpenRouter pricing, checked on . Lowest paid plan: Standard: pay-as-you-go with no minimum; a 5.5% fee on credit purchases. Free tier: Free: 25+ free models from 4 free providers, 50 requests/day, rising to 1,000 a day once the account has bought at least $10 of credits. Pricing model: Prepaid credits spent at each provider's list price, plus a platform fee on credit purchases: 5.5% on Standard, 8% on Business.

Plan Price Included usage Beyond that Data retention Users
Free Free 25+ free models, 4 free providers, 50 requests/day – – –
Standard Pay-as-you-go, 5.5% platform fee on credit purchases 500+ models, 80+ providers – – –
Business Pay-as-you-go, 8% platform fee on credit purchases 500+ models, 80+ providers – – –
Enterprise Set by contract; fee discounts available 500+ models, 80+ providers – – –

There is no flat plan price: the platform fee is charged when credits are bought, not on individual requests, and inference is billed at the provider's list price on every plan. Standard and Business have no minimums and no lock-in. With your own provider keys (BYOK), Standard and Business include "$25,000 of list price inference / month with no fees, 5% fee after". Rate limits on paid plans are "Passthrough from the provider(s)". Workspaces are limited to 5 on Free and Standard and 1,000 on Business; the page states no user limit. Enterprise fees are set by contract, with discounts on OpenRouter's fees but not on inference. The documentation FAQ, not this page, adds a $0.80 minimum fee on card purchases, a 5% fee on crypto payments, a 1% discount on usage for opting in to OpenRouter's use of inputs and outputs, and that unused credits may expire one year after purchase.

As stated on the vendor's pricing page on . A dash means the page does not say.

OpenRouter pricing history

The pricing page has been checked daily since . No price change has been recorded since then.

Where OpenRouter fits

Teams that want to call models from many providers with one key and one bill, without a contract with each provider and without running a gateway. There is no monthly fee or minimum on the Standard plan, and the free plan allows 50 requests a day to free models for trying it out.

Limits of OpenRouter

Every credit purchase carries a fee, 5.5% on Standard and 8% on Business, on top of the providers' list prices. It cannot be self-hosted, and EU or US in-region routing needs the Business plan or above. Prompts are not stored unless logging is switched on, so it is not a tracing tool by default. Unused credits may expire a year after purchase.

Other gateways and observability tools

  • Langfuse

    Langfuse is an open-source platform for tracing, evaluating and managing prompts in applications built on language models.

  • LiteLLM

    LiteLLM is an open-source Python SDK and self-hosted gateway that calls many language model providers through the OpenAI API format.

COMPARE THEM SIDE BY SIDE →