LongCat announced that LongCat-Flash-Thinking is now available through the platform API and compatible with the Claude Code workflow. The platform has also increased its daily free quota from 100,000 tokens to 500,000 tokens/day . Users can submit an application on the Usage page and, upon approval, receive a higher quota of 5,000,000 tokens/day . To support this upgrade, the official API documentation, Quick Start, and FAQ have been updated, and a change log and Claude Code configuration guide have been added to facilitate rapid developer integration and migration.
According to the platform's instructions, quotas automatically refresh daily at midnight (Beijing time) and do not carry over. Higher quotas require submission of usage and compliance information according to the guidelines. Model weights and sample code are maintained simultaneously on GitHub and Hugging Face, facilitating verification in local or third-party inference environments. The community is also accelerating the adaptation of inference service providers and toolchains.
Frequently Asked Questions
Q: How to use LongCat-Flash-Thinking API?
A: Register on the LongCat platform and obtain a key, then call it according to the Quick Start configuration in the API Docs. It supports integration with Claude Code.
Q: What is the free quota? Can it be increased?
A: The default is 500,000 tokens/day ; you can apply for up to 5 million tokens/day on the Usage page, which needs to be approved by the platform.
Q: How are quotas settled and refreshed?
A: Statistics are calculated daily and automatically refreshed at midnight (Beijing time zone). Unused funds will not be carried forward. Usage can be viewed at /status .
Q: Where can I find updates and configuration examples?
A: Please refer to the Change Log and Claude Code configuration documentation. The API Docs, Quick Start, and FAQ have been updated synchronously.
Q: Are there any open source resources for reference?
A: Yes. Models and examples are available on GitHub and Hugging Face for easy local verification and comparative evaluation.
