DeepSeek API Reopens: Access Restored After Pause

DeepSeek Restores API Access After Temporary Suspension

The Chinese artificial intelligence company, DeepSeek, has once again made its API available to users. Access had been temporarily suspended for almost three weeks because of limitations in available resources.

Beginning on Tuesday, customers were permitted to add credits to their accounts for utilizing the DeepSeek API. This API enables developers to create applications and services leveraging cloud-based versions of DeepSeek’s AI models.

However, a company spokesperson communicated via WeChat, as reported by Bloomberg, that server capacity continues to be under pressure during peak hours.

Rise of DeepSeek and its R1 Model

DeepSeek gained significant attention earlier in the year with the release of its R1 model, designed for advanced “reasoning” tasks. Performance benchmarks indicate that R1 either equals or surpasses that of several of OpenAI’s leading AI models.

This level of competitiveness from DeepSeek has reportedly led OpenAI to evaluate the possibility of open-sourcing more of its own technology and adjusting the timing of certain product launches.

Increased Competition in the Chinese AI Landscape

Alongside DeepSeek, other Chinese technology companies are actively increasing their capacity for AI model production.

Notably, Alibaba, a major Chinese tech firm, simultaneously launched a preview of its newest reasoning AI model, QwQ-Max, on the same day DeepSeek resumed API credit top-ups. Alibaba intends to make QwQ-Max open source.

Here's a summary of the key developments: