Deepseek has unveiled its V4.1-Flash-1017 model, which has shown remarkable improvements over its predecessor in agentic tasks. This open-weight model not only achieves better benchmark results but also completes tasks more quickly and generates shorter outputs, making it a competitive option against larger rivals in the AI landscape.
In addition to the model release, Deepseek is launching new Search and Fetch APIs, priced at $0.75 per 1,000 calls. This move aims to enhance accessibility for developers, particularly those utilizing tools like Claude Code and Codex, which currently enjoy free access to these APIs. This strategic decision is expected to broaden Deepseek’s user base and facilitate more efficient AI integration into various applications.
The advancements in V4.1-Flash-1017 and the introduction of the new APIs signify Deepseek’s commitment to innovation and cost-effectiveness in the AI domain, positioning the company as a formidable player in the market.
Compiled automatically by the Tech AI Newsdesk from public AI-news sources and summarised in our own words.
