ZotGPT API Now Available for Research

Oct 17, 2024 | Home Page News, News

Application programming interface (API) access for ZotGPT is now available for all faculty and staff. This highly anticipated release supports UCI research by providing access to ZotGPT’s gen AI capabilities securely and privately. ZotGPT safeguards the proprietary data of UCI affiliates and prevents companies like OpenAI from training their models on our data.  

Getting Started:

  1. Request an API Key: Your request will be reviewed and approved, then you’ll receive an API key via email. This key will enable access to the ZotGPT service.
  2. Provide Recharge Information: After obtaining the API key, complete the recharge information by submitting a ServiceNow request form. This step is necessary to facilitate any subsequent usage charges.

Additional Details:

To help UCI faculty and staff get started, OIT is offering each individual a $200 credit for API token usage. Thereafter, cost will be recharged back to the KFS information provided. 

We encourage our UCI community to explore the capabilities of ZotGPT and integrate it into their projects. For any questions or assistance with the setup process, please do not hesitate to contact our support team at oit@uci.edu.

Article updated on:
February 7, 2025

0 Comments

More News from OIT

Claude 3.7 Now Available in ZotGPT Chat

UC Irvine has successfully implemented Claude 3.7 into its ZotGPT Chat platform. The integration of Anthropic’s newest AI model occurred within hours of its public release, demonstrating UCI's commitment to providing up-to-date AI capabilities to its community. Claude...

IT Service Impacts

Due to several planned strikes involving UC workers this week, some of our OIT services will be impacted and may result in longer service times and delays to your IT requests.

Duo Verified Push Coming to UCPath

OIT is enhancing UCPath security with Duo “Verified” Push. This feature requires entering a unique code displayed on the login screen, adding protection against phishing and accidental approvals when suspicious activity is detected.