Last updated: 13 November 2026. All prices checked on the vendors’ official pages that day.
OpenRouter is a single API that routes requests to hundreds of models from many providers, including OpenAI, Anthropic, Google, Mistral and DeepSeek. People search for OpenRouter pricing to answer two questions: does it cost more than going direct, and what are the limits on free models? This guide answers both with figures from OpenRouter’s pricing page and public models API on 13 November 2026, then shows how to create and safely cap an OpenRouter API key.
What is OpenRouter?
OpenRouter gives you one endpoint (https://openrouter.ai/api/v1) and one key for 600+ models from 80+ providers on paid plans. It can fall back to another provider when one fails and offers an auto-router that picks a model for each request. Its API accepts OpenAI-style SDKs by changing the base URL.
OpenRouter pricing: plans and fees
| Plan | Platform fee | Models / providers | Key extras |
|---|---|---|---|
| Free | None | 25+ free models, 4 free providers | 50 requests/day, community support |
| Standard | 5.5% on credit purchases | 600+ models, 80+ providers | Budgets, prompt caching, auto-routing, email support |
| Business | 8% on credit purchases | 600+ / 80+ | EU/US in-region routing, 1,000 workspaces |
| Enterprise | Contract (fee discounts available) | 600+ / 80+ | Invoicing, SSO/SAML, contractual SLAs |

- No markup on inference. OpenRouter says inference is billed at the provider’s list price on every plan. Its fee is charged when you buy credits (5.5% on Standard, 8% on Business), never per request.
- BYOK (bring your own key). If you use your own provider keys through OpenRouter, the provider bills you directly. Standard and Business include $25,000 of list-price inference per month through BYOK with no OpenRouter fee; above that, OpenRouter charges 5% of the list-price cost.
- Failed attempts are not billed for model tokens, so with fallback you pay only for the attempt that answers. Routing through
openrouter/autoadds no fee. - Payment: cards, AliPay and crypto (USDC) on Standard and Business; invoicing and purchase orders on Enterprise. Prices exclude VAT/GST where applicable. No minimum spend or lock-in on Standard.
Free models and rate limits
Free models are capped at 20 requests per minute and 50 requests per day, rising to 1,000 per day once your account has bought at least $10 of credits. Failed attempts count toward the daily cap, and a negative balance blocks free models. On paid models OpenRouter sets no plan-based rate limit; limits come from the underlying providers. Free models change often, so do not build production features on them.
Model prices: OpenRouter vs going direct
| Model on OpenRouter | Input / 1M | Output / 1M | Direct list price |
|---|---|---|---|
| openai/gpt-6.1-sol | $2.00 | $10.00 | $2.00 / $10.00 (OpenAI) |
| anthropic/claude-sonnet-5.5 | $2.00 | $10.00 | $2.00 / $10.00 (Anthropic) |
| google/gemini-4-argon | $2.00 | $10.00 | $2.00 / $10.00 until 30 Nov (Google) |
| mistralai/mistral-large-4-0 | $0.70 | $2.10 | $0.70 / $2.10 (Mistral) |
| google/gemini-3.8-flash | $0.825 | $4.125 | $0.75 / $3.75 (Google) |

For most flagship models the per-token price on OpenRouter matched the vendor’s own list price on the day we checked. Some routes differ (Gemini 3.8 Flash showed $0.825 / $4.125 on OpenRouter versus Google’s $0.75 / $3.75 promotional list price), so always check the model page. Your real extra cost is the credit-purchase fee, plus any vendor-specific discounts you might not get through a router. For direct vendor prices, see our LLM API pricing comparison, OpenAI pricing, Claude pricing and Gemini pricing.
How to create and cap an OpenRouter API key

- Create an account (use a company email; create an organisation if several people need access).
- Buy credits. The Standard fee is added to the purchase.
- Create an API key in your account settings. Give each app or environment its own key.
- Set a credit limit on the key, optionally with a daily, weekly or monthly reset. OpenRouter recommends hard caps for keys used by agents, CI jobs and service accounts.
- Call the API with your key as a Bearer token.
- Review activity logs and exports regularly.
export OPENROUTER_API_KEY="your_openrouter_key"
curl https://openrouter.ai/api/v1/responses \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "openai/gpt-6.1-terra", "input": "Hello"}'Teams that automate key creation can use Management API keys, which can also set expiry dates: once a key expires, requests fail with 401 API key expired.
When OpenRouter makes sense, and when it does not
- Good fit: prototyping across many models, fallback between providers, small teams that want one bill.
- Think twice: workloads with strict data-processing or residency requirements (in-region routing is Business/Enterprise only), or where you need vendor-specific features first.
- Not a workaround. An aggregator does not change who may use a model. Each provider’s policies and OpenRouter’s terms still apply, so do not treat it as a way around regional restrictions or KYC.
Some companies prefer direct accounts with each vendor; others want one unified API with consolidated usage reporting. We compare the two in Token as a Service for business. PandaWM helps companies open their own official vendor accounts or connect through a compliant unified API: Overseas AI Token service: account setup + API access (PandaWM).
FAQ
Is OpenRouter free?
There is a Free plan with 25+ free models and a cap of 50 requests per day (1,000 after buying $10 of credits). Paid models cost the provider list price plus a fee on credit purchases.
Does OpenRouter charge more than OpenAI or Anthropic?
It says it does not mark up inference; the cost difference is the 5.5% (Standard) or 8% (Business) fee on credit purchases, and some routes are priced differently.
What is the OpenRouter BYOK fee?
Free up to $25,000 of list-price inference per month on Standard and Business, then 5%.
Are failed requests billed?
No model-token charge for attempts that error. Auxiliary services that already ran may still be billed.
Can I limit how much an OpenRouter key spends?
Yes. Set a credit limit per key with optional periodic reset.
Is PandaWM an OpenRouter partner?
No. We mention OpenRouter as one option and do not resell it.
Compliance note: Each model vendor only serves customers in its supported countries and regions and requires accurate account, identity and billing information. PandaWM helps companies use AI models through their own accounts or overseas entities that are legitimately established in supported regions, or through a compliant unified API. We do not sell, rent, share or transfer accounts or API keys, and we do not help anyone bypass regional restrictions, KYC, sanctions or payment controls. We are not an official partner or reseller of OpenAI, Anthropic, Google or any other vendor. Prices on this page are the vendors’ public list prices, not PandaWM prices.
Related guides
- Overseas AI Token service: account setup + API access (PandaWM)
- LLM API pricing comparison: cheapest providers
- OpenAI API pricing (GPT-6) + cost calculator
- Claude API pricing: Opus, Sonnet, Haiku
- Gemini API pricing, free tier and billing tiers
- How to get an OpenAI API key
- Token as a Service (TaaS) for business
- 简体中文:大模型 API 价格对比
Talk to PandaWM about overseas AI Token access
Tell us which models you need (GPT, Claude, Gemini or others), your expected monthly token volume, where your company and any overseas entity are registered, and whether you prefer your own vendor accounts or one unified API. We help companies open and manage their own official accounts (billing, keys, usage controls) or connect through a compliant unified API. We do not sell, rent or share accounts, and we are not an official reseller of any model vendor.
WhatsApp: +852 5904 8602Telegram: @PANDA0134WeChat: szcwh99

Official sources
All official pages checked on 2026-11-13. Prices are in USD per 1 million tokens (MTok), before tax. Vendors change prices; the official pages prevail.

微信扫一扫打赏
支付宝扫一扫打赏
