Never Pay for Empty AI Responses Again
Today we're excited to announce the expansion of our Zero Completion Token Insurance coverage to every model available on OpenRouter! Starting now, if you receive a response with no output tokens—due to an empty or error finish reason—you will not be charged, even if the underlying model provider charges us for prompt processing.
At today's usage volumes, this initiative is projected to save our users over $18,000 per week across hundreds of models, reinforcing our commitment to fairness and reliability.
How It Works
Zero Completion Insurance automatically protects your account when:
- A response has zero completion tokens with a blank or null finish reason.
- A response has an error finish reason.
In either scenario, no credits will be deducted from your account, regardless of provider charges to OpenRouter.
Automatic and Hassle-Free
Zero Completion Insurance is automatically enabled for every OpenRouter user. No additional configuration or action is needed from your side.
Viewing Protected Requests
Protected requests can easily be identified on your Activity page - any request with 0 output tokens will also have a 0 for the cost.
Our Commitment
As the first and largest LLM router, we've encountered virtually every possible quality issue from model providers. We're dedicated to improving quality standards industry-wide and will proactively collaborate with all providers behind the scenes. Our goal is to create a sustainable and user-friendly ecosystem.
Stay tuned for more enhancements as we continually strive to improve the inference experience for everyone. Your feedback matters greatly to us—please reach out directly or join the conversation in our Discord community: openrouter.ai/discord.
Thank you for your continued support!