Deepseek released Deepseek-V4.1-Flash-1017 through its website, app and API, and published the model weights as open source on Hugging Face and ModelScope. The company says the model was trained with reinforcement learning using self-feedback, and that it runs faster and produces shorter outputs than its predecessor.
In Deepseek's internal benchmarks, the model beats Claude Opus 5, GPT 5.6-Sol, Kimi K3 and GLM 5.3 in several coding and agent categories. A cost chart shows a top DeepSWE score of 77.0 percent at an average of just under $0.20 per task. Independent benchmarks have not yet been published.
Deepseek is also making its Search and Fetch APIs available to all developers. The company says they run on an internal search engine that handles hundreds of millions of queries per day. API users can use the features free through server-side tools in Deepseek Harness, Claude Code and Codex until October 31, 2026.
Why it matters if you build agents
Developers building coding and agent workflows can now call Deepseek's search and fetch endpoints directly, at a stated price of $0.75 per 1,000 calls. The free access period through supported tools ends on October 31, 2026, so cost planning should account for it.
Treat the cost and benchmark claims as company figures until independent tests are available.
Get stories like this every weekday
One short email with the day's agent tooling news.