Get the support you need to succeed with OpenRouter. Browse our help center, raise a ticket, or explore our documentation to find answers quickly.
Browse our comprehensive guides and documentation
Get personalized help from our support team
Technical documentation and API references
Find answers to common questions about OpenRouter
OpenRouter provides a unified API to access major LLM models, aggregate billing, and track usage with analytics.
We pass through provider pricing while pooling uptime, so you get the same provider rates with a unified API and automatic fallbacks for better reliability.
Create an account and add credits on the Credits page. Credits are used for LLM inference.
Use the chat room or create API keys and start using the API. See our quickstart guide.
Join our community Discord and ask in the #help forum.
For billing and account questions, email [email protected].
Pricing is per model and may differ for prompt/completion, images, and reasoning tokens. See the model page for details.
We compute cost from provider token usage and deduct it from your credits. Review your history in Activity.
Include usage: {include: true} in requests to receive usage data in the response. We pass through provider pricing; we charge a small fee when purchasing credits.
We charge a small fee when you purchase credits. Model inference pricing is passed through from providers without markup.
Crypto payments may include an additional processing fee.
Using BYOK includes a free monthly request allowance; subsequent usage incurs a small percentage fee based on the equivalent OpenRouter cost. This is deducted from your credits.
Variants are suffixes that modify behavior. Static variants are available on specific models, listed in the models API:
:free – free model with low rate limits:beta – not moderated by OpenRouter:extended – longer context length:thinking – reasoning by defaultDynamic variants can be used on all models:
:online – attach web results to the prompt:nitro – optimize for faster throughput:floor – prioritize lowest price:nitro to route to faster providers.Free model rate limits depend on your purchased credits. See details in the rate limits documentation.
/completions and /chat/completions, plus /api/v1/models. See the API docs.stream: true.See our Terms of Service and Privacy Policy.
We log basic request metadata (timestamps, model, token counts). Prompts and completions are not logged by default.
You can opt in to log prompts/completions for a small discount via privacy settings.
For Stripe purchases, allow up to one hour. If no charge/receipt appears, your card may have been declined; try a different method.
If charged but credits are missing, email [email protected] with purchase details. For crypto payments, email support as well.
Go to Settings → Manage Account → Security tab to delete your account.
Note: unused credits will be lost and cannot be reclaimed if you delete and later recreate your account.