OpenRouter pricing (2026): is OpenRouter free, and what does the fee cost you?
updated ยท prices checked on vendor pages
quick answer: OpenRouter charges each model's list price per token, with no markup. its cut is a fee when you buy credits: 5.5% on Standard (pay as you go), 8% on Business, discounts on Enterprise. there's a Free plan with 25+ free models capped at 50 requests a day, and bring your own key (BYOK) is fee-free for the first $25,000 of list-price inference a month.
OpenRouter is #4 of 10 on our LLM APIs & models board (8.5/10) because it's the easiest way to try lots of models from your own code: one key, one OpenAI-compatible API, and swapping models is a one-line change. you don't pick it for cheap tokens. you pick it so you're not married to one provider.
the plans
as of oct 30, 2026, from OpenRouter's pricing page:
| plan | fee | models | worth knowing |
|---|---|---|---|
| Free | none | 25+ free models from 4 providers | 50 requests a day, no budgets, no prompt caching, community support |
| Standard (pay as you go) | 5.5% on credit purchases | 600+ models, 80+ providers | budgets and spend controls, auto-routing, prompt caching, 5 workspaces, email support |
| Business (pay as you go) | 8% on credit purchases | 600+ models, 80+ providers | adds EU and US in-region routing, advanced notifications, workload identity federation, 1,000 workspaces |
| Enterprise | discounts on the fee | 600+ models, 80+ providers | SSO/SAML, contractual SLAs, invoicing, a dedicated support team |
no minimum spend and no lock-in on Standard or Business. you can upgrade to Business from your settings without talking to sales, and downgrade 30 days later.
is OpenRouter free?
partly. the Free plan gets you chat and API access to 25+ free models with no card. the limits are tight on purpose:
- 20 requests a minute and 50 a day on free models.
- 1,000 a day once your account has bought at least $10 of credits, so one small top-up lifts the cap a lot.
- failed attempts count toward that daily cap, and a negative balance blocks free models until you top up.
fine for kicking the tyres or a hobby bot. not something to put a real app on.
what OpenRouter really costs
- tokens are list price. OpenRouter says inference is billed at the provider's list price on every plan, and it doesn't discount inference either. streaming costs the same.
- the fee is on the credits, not the requests. buy $100 of credits on Standard and the fee is 5.5% of that, charged when you buy. on Business it's 8%. the only billing difference between the two is that fee.
- failed requests are free. an attempt that errors isn't charged for its model tokens, so with fallback you pay for the one that answers. extras that already ran, like web search or file parsing, may still be billed.
- auto-routing adds no fee.
openrouter/autocharges the rate of whichever model serves the request. - prices can change under you. if a model's price changes, you're charged the new rate. pin an explicit model ID if you don't want surprises.
- taxes are extra where they apply.
BYOK: bring your own key
already have an Anthropic, OpenAI or Google key? on Standard and Business the provider bills you directly, and OpenRouter waives its fee on the first $25,000 of list-price inference a month. above that it charges 5% of what the request would have cost on OpenRouter, taken from your credits. on Enterprise the allowance is custom. so for most indie hackers, BYOK through OpenRouter costs nothing extra and buys you the routing and fallback.
is the fee worth it for a dev?
- you use one model, all day: go direct. the Anthropic Claude API is #1 (9.0/10) for coding and agents, and skipping the middleman skips the 5.5%.
- you're comparing models or want fallback when a provider falls over: OpenRouter. 5.5% is a fair price for not writing your own router.
- you already have provider keys: BYOK on Standard. fee-free up to $25,000 a month.
- you need data kept in the EU or US: Business, for the in-region routing (send requests to eu.openrouter.ai or us.openrouter.ai).
OpenRouter also says it doesn't train on your data, and zero data retention routing is on every plan.
OpenRouter vs going direct
- Claude direct: Anthropic Claude API, #1 on our board, the one to beat for coding agents
- the value pick with a free tier: Google Gemini API, #2 (8.5/10). see Gemini API pricing
- the biggest ecosystem: OpenAI API, #3 (8.5/10)
- open models on your own machine, free: Ollama. see Ollama pricing
the rest of the field
the table: the rest of the field
| rank | tool | best for | free tier | paid from | numptr score |
|---|---|---|---|---|---|
| #4 | OpenRouter | prototyping across providers | Free plan available | pay-as-you-go at model list prices, 5.5% fee on credit purchases | 8.5/10 |
| #1 | Anthropic Claude API | coding agents, tool use | small amount of free credits for new API users | Sonnet 5.5 $2/$10 per M tokens | 9.0/10 |
| #2 | Google Gemini API | long-context and agentic tasks on a budget | free tier on Flash models (Argon not on free tier) | Gemini 4 Argon $2/$10 per M tokens through nov 30, 2026, then $4/$20 | 8.5/10 |
| #3 | OpenAI API | general-purpose and multimodal apps | unknown | GPT-6.1 Sol $2/$10 per M tokens | 8.5/10 |
faq
is OpenRouter free?
there's a Free plan with 25+ free models, capped at 20 requests a minute and 50 a day (1,000 a day once you've bought $10 of credits). paid models need credits.
does OpenRouter mark up model prices?
no. tokens are billed at the provider's list price on every plan. OpenRouter's fee is 5.5% on Standard or 8% on Business, charged when you buy credits.
what's the difference between OpenRouter Standard and Business?
the Business plan costs 8% on credit purchases instead of 5.5%, and adds EU and US in-region routing, advanced notifications, workload identity federation and up to 1,000 workspaces.
does BYOK cost extra on OpenRouter?
not for the first $25,000 of list-price inference a month on Standard and Business. after that, OpenRouter charges 5% of the list price, deducted from your credits.
does OpenRouter have rate limits?
no plan-based limits on paid models: limits come from the provider. free models are capped at 20 requests a minute and 50 or 1,000 a day.
tools in this guide
sources
checked 2026-10-30
- OpenRouter pricing (Free, Standard 5.5%, Business 8%, Enterprise fee discounts, model and provider counts, plan features, workspaces, no markup on list price, fee on credit purchases, BYOK $25,000/month then 5%, free model limits 20/min and 50 or 1,000 a day, failed attempts not billed, openrouter/auto no fee, no minimums, Business upgrade and 30-day downgrade, in-region routing, no training on data, zero data retention on every plan, taxes): https://openrouter.ai/pricing
- ranks, scores, verdicts and plans: numptr data/tools.json (LLM APIs & models board)