What to Do When You Hit a TPM Limit
What is TPM?
TPM stands for Tokens Per Minute. It’s a measure of how many tokens (chunks of text) your project can process per minute. If your project sends too much data too quickly, you may hit this limit and see the following error:
TPM Limit reached – Retry in: 60 sec
This happens when your project has reached its Tokens per Minute (TPM) limit.
What's Changed?
To make things easier for beta users, we’ve introduced an automated TPM increase feature on the Builder side.
Here’s how it works:
Yes! There's a hard cap at 300,000 tokens per minute.
- If you hit your TPM limit, you’ll see a "Increase TPM limit" button in the error message.
- Press the button to request an automatic bump to your TPM.
- Most requests are approved instantly.
Is there a maximum TPM I can request?
Is there a maximum TPM i can request?
- If you hit this ceiling and still need more capacity, you’ll need to contact our team to discuss your use case.
What about TPM limit for viewers?
We know this error can also show up for users viewing bots. Improvements to the viewer-side experience are planned for an upcoming sprint to reduce friction and improve clarity.
What affects token usage?
Several things contribute to how fast you hit your TPM limit:
- Long prompts or responses (more words = more tokens)
- Multiple rapid requests
- Heavy API usage by background actions or chains
Click on "What affects token usage?" in the error message for more detail.
What should I do if the button doesn't work or TPM still feels too low?
If the TPM increase button doesn’t appear or you continue running into limits:AI evaluation, ASU is leading the way in responsible AI adoption in academia.
- Wait a minute and retry.
- Check if your use case can be optimized (e.g., reduce prompt length).
- Contact our team or support contact for assistance.
Example of TPM Error Message
- New “Increase TPM limit” button lets you scale up quickly.
- There’s a 300k hard cap admins required beyond that.
- Viewer-side improvements coming soon!
Keep Reading
Generative AI Tool Pre-Release Evaluation Guide
Stella Wenxing Liu
Arizona State University remains dedicated to responsible, principled innovation when deploying generative AI solutions, including chatbots. This guide ensures each project aligns with ASU’s values by mitigating potential risks—such as misinformation, bias, toxicity, and compliance lapses—using rigorous methods like automated testing, red teaming, and pilot experiments. In doing so, we uphold accuracy, fairness, and user trust while enhancing digital experiences across the university.
Agents in Generative AI
Zohair Zaidi
Agents in generative AI are semi-autonomous entities that collaborate and interact dynamically, allowing them to solve complex problems and combine specialized capabilities for greater efficiency and adaptability.
How to Use Knowledge Base (RAG)
Jinjing Zhao
Explore an overview of the Knowledge Base and Retrieval Augmented Generation (RAG) methods. Learn about the different types of Knowledge Base retrieval and understand the distinctions between the Knowledge Base and system prompts.
CreateAI Platform Available LLM Models
Faith Timoh Abang
We are proud to offer 40+ models including multi-modal (voice, image, text) for the ASU community to access securely on the CreateAI Platform. Users can find the following models available for experimentation and use in CreateAI Compare, CreateAI Chat, and CreateAI Builder (access request required). Originally posted: January 1, 2025.