Buying snapshot
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Decision point | What OpenRouter states |
|---|---|
Getting started | Create an account, buy credits usable with any model or provider, and create an API key. Switching an OpenAI-based app means updating the base URL and model names. |
Model charges | Input and output tokens are billed per model at posted rates, at the provider's list price on every plan; streaming does not change the per-token price. |
Commitment | Standard and Business: no minimums and no lock-in; you pay only for what you use. |
Payment | Standard and Business: credit and debit cards, AliPay, and crypto (USDC). Prices exclude applicable taxes; VAT/GST may be added on invoices where required. |
Support | Enterprise: support SLA with committed response times, dedicated Slack channel, customer success manager and a forward-deployed engineer. Standard and Business: email support. Every plan: community Discord. |
Our take
If you want to reach many providers' models through one OpenAI-compatible API and pay per token with no minimums, OpenRouter is a straightforward choice, and the self-serve Standard plan covers that. If you need SSO, SCIM or lower OpenRouter fees, budget for an Enterprise contract instead.
Evaluate OpenRouter
Comparisons
Where it stands out
Provider fallback without paying twice. OpenRouter can fall back to other providers when one goes down, and an attempt that errors is not charged for its model tokens, so you pay only for the run that answers. Under Zero Completion Insurance the same holds for an empty response with no finish reason; auxiliary services that already ran, such as web search or file parsing, may still be billed.
No plan-based rate limit on paid models. Limits there normally come from the provider, and fallback routing retries other providers automatically. Only very high burst traffic is rejected by OpenRouter's DDoS protection with a 429.
Control over where prompts go. On Standard, Business and Enterprise, data policy-based routing with model and provider policies keeps prompts to the models and providers you trust. OpenRouter says it does not train on your data; provider-side retention can be turned off for the account or per request, and Zero Data Retention routing is available on every plan.
Switching models is cheap. You can switch models without changing your integration. On Standard, Business and Enterprise, auto-routing is available, and routing through openrouter/auto adds no fee: you pay the rate of whichever model serves the request.
Product screenshot
Trade-offs to weigh
The fee grows with spend. OpenRouter's fee is a percentage of each credit purchase, so it rises with usage. Discounted OpenRouter fees are offered only on Enterprise, and OpenRouter does not discount inference itself.
Model changes reach you. If a model's price changes, OpenRouter keeps routing to it and charges the new rate; a deprecated model returns a 404. Choose an explicit model ID/version to avoid surprises.
Latency is not fixed. Routing improves reliability, but latency varies by model, provider and region. For consistent latency, pin a specific model and provider, which gives up some of the routing benefit.
Identity controls need Enterprise. SSO (SAML) and SCIM are available on Enterprise plans only after you contact sales, so a team with a single sign-on requirement cannot self-serve on Business.
Pricing and plan conditions
On Standard and Business you buy credits, manually or with auto top-up, and each request deducts its cost from the balance. The fee is charged when you buy credits, never on individual requests.
Bring your own key (BYOK). The provider bills you for inference directly. Standard and Business include $25,000 of list-price inference per month through BYOK with no OpenRouter fee; Enterprise allowances are custom. Above the allowance, OpenRouter charges 5% of what the same request would cost at list price on OpenRouter, deducted from your credits.
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Plan | Platform fee | What it adds |
|---|---|---|
Free | N/A | 25+ free models from 4 free providers, chat and API access, 5 workspaces, community support. Free models: 20 requests per minute and 50 per day, rising to 1,000 per day once the account has bought at least $10 of credits; failed attempts count toward the daily cap, and a negative balance blocks free models. |
Standard (Pay-as-you-go) | 5.5% on credit purchases | 600+ models from 80+ providers, 5 workspaces, email support. |
Business | 8% on credit purchases | 1,000 workspaces, EU and US in-region routing (eu.openrouter.ai, us.openrouter.ai), Workload Identity Federation. Upgrade from settings with no contract, no minimum spend and no sales conversation; you can downgrade 30 days after the upgrade. |
Enterprise | Fee discounts available | Contract based on volume, prepaid credits and annual commitments; invoicing and purchase orders through sales. In-region routing, Workload Identity Federation, custom BYOK allowances. Private Endpoints are an add-on. |
Public signals
About these signals
Verified factual signals from public sources. They indicate observable activity or interest, not total adoption, product quality, or cost.
509.6k PyPI weekly downloads4.3M npm weekly downloads0 vulnerabilities across 2 packages
See all signals from 4 sources
npm · @openrouter/ai-sdk-provider@3.1.0 · PyPI · openrouter@1.3.144
Alternatives to OpenRouter
The reviewed substitutes for OpenRouter among the LLM gateways, and what would make each one the better answer.
Direct alternatives
Reviewed substitutes: products bought for the same job, where a team picks one.
- LiteLLM
- Both are LLM gateways putting one API in front of many providers. LiteLLM is self-hosted and OpenRouter is a hosted service; a team routes through one of them.