This repository offers a consolidated, community-driven summary of knowledge cut-off dates for numerous large language models (LLMs) across major providers like OpenAI, Google, and Anthropic. It serves as a vital reference for developers and researchers to understand the temporal scope of LLM knowledge, crucial for applications requiring up-to-date information.

How It Works

The project aggregates LLM knowledge cut-off data from official technical reports, API documentation, and public discussions. Information is meticulously compiled and presented in a clear, tabular format, allowing for straightforward comparison of temporal data limitations across different model families and versions.

Quick Start & Requirements

This repository serves as a data reference and does not require installation or execution.

Highlighted Details

• Comprehensive coverage of models from OpenAI, Google, Anthropic, Meta, Qwen, DeepSeek, and Microsoft.
• Includes a broad spectrum of models, from historical versions to the latest previews and future-indicated releases (e.g., GPT-5, Claude 4).
• Actively maintained with recent updates and a call for ongoing community contributions to ensure accuracy and completeness.
• Details specific cut-off dates where available, clearly marking entries as "Unknown" or "TBD" when information is not publicly disclosed.

Maintenance & Community

The project relies on community contributions for updates and additions, with an explicit invitation for users to help maintain the accuracy of LLM knowledge cut-off dates. No specific community channels or contributor details are listed.

Licensing & Compatibility

No license information is provided within the README, which may pose compatibility concerns for commercial or integrated use.

Limitations & Caveats

Knowledge cut-off dates for several models, including those from Mistral AI and xAI (Grok), are listed as unknown. The accuracy and currency of the data are contingent upon the availability of public information and the diligence of community contributors. Placeholder "Source" links are present instead of direct URLs.