DeepSeek Launches V4.1-Flash-1017, Scores 77.0 on DeepSWE and Opens Search API to Developers
DeepSeek has released DeepSeek-V4.1-Flash-1017, a new checkpoint of its V4.1-Flash model. The update is live on the DeepSeek website, app, and API. DeepSeek claims stronger benchmark scores, faster task completion, and shorter outputs than the September V4.1-Flash release, positioning the update as a Pareto improvement in both performance and cost.
Alongside the model, DeepSeek has opened its Search and Fetch APIs to every developer. For agent builders in 2026, this means a cheaper coding model plus a first-party web layer behind a single API key—a combination that reduces the need to integrate third-party search providers into agent pipelines.
TL;DR
- Size: 552B backbone parameters plus 196B Engram parameters. The model activates 8B parameters per token during prefill and 16B during decode. Context is 1M tokens, with up to 384K output tokens.
- Runs on: DeepSeek API as
deepseek-flash. The September V4.1-Flash weights (763B parameters, FP8) are MIT-licensed on Hugging Face. Hardware requirements and the 1017 weights have not been disclosed. - Performance: Beats the previous V4.1-Flash on all 11 benchmarks DeepSeek reported, by 2.0 to 11.1 points.
- Best: 77.0 on DeepSWE v1.1, above Claude Opus 5 (74.0) in DeepSeek's table.
- Worst: 42.6 on HLE without tools, versus 56.3 for Claude Opus 5.
via MarkTechPost