DeepSeek V4.1 Flash Gets New Model With Big Coding Gains, Cheaper Agent Tasks

Chinese AI firm DeepSeek has issued DeepSeek-V4.1-Flash-1017, an updated version of its Flash model with improvements in coding, agentic tasks, speed, and output efficiency.

Key points

• On the DeepSWE benchmark, V4.1-Flash-1017 reached a top score of 77.0% at an average cost of just under $0.20 per task.
• DeepSeek and Moonshot Accused of Secretly Routing Over 300, 000 Users to Claude.
• DeepSeek API users can access the tools for free through server-side integrations in DeepSeek Harness, Claude Code and Codex until October 31, 2026, at 4: 00 PM UTC.
• DeepSeek tuned V4.1-Flash-1017 for development environments including DeepSeek Harness, Tencent CodeBuddy and WorkBuddy.
• According to DeepSeek's own benchmarks, V4.1-Flash-1017 improves over the previous V4.1-Flash throughout every tested category.

Notably, the model is available through DeepSeek's website, app, and API. The firm has additionally published its weights on Hugging Face and ModelScope.

DeepSeek notes the update delivers better benchmark performance than its predecessor while completing tasks faster and generating shorter responses. The firm describes this as a Pareto improvement in the balance between performance and cost. DeepSeek V4.1 At present 1500x Cheaper Than Usual.

For end-to-end software development, DeepSeek uses a post-training method called reinforcement learning with self-feedback, or RLSF.

For context, the technique allows the model to employ its own confidence scores or generated critiques as reward signals rather than relying on human evaluators or external reward models.

DeepSeek notes this improves the model's ability to build complete projects from scratch, covering tasks including game development, 3D modeling, and design work. The model can additionally create compact animated explainer videos.

DeepSeek tuned V4.1-Flash-1017 for development environments including DeepSeek Harness, Tencent CodeBuddy and WorkBuddy. The firm recommends DeepSeek Harness version v0.2.1-rc.1 or newer. Smaller Model Beats Larger Rivals.

For context, the firm additionally claims the smaller model outperforms larger systems including Claude Opus 5, GPT-5.6 Sol, Kimi K3 and GLM 5.3 in a number of coding and AI agent benchmarks.

At its low setting, the model still scored above 72% while costing around $0.10 per task.

Its predecessor reached 74.2% but cost roughly $0.50 per run.

While showing lower accuracy than V4.1-Flash-1017, deepSeek's chart places models including GPT-6 Astra, Gemini 3.8 Flash and Claude Opus 5 at between roughly $2 and more than $10 per task.

Independent benchmarks are not yet available, and real-world developer assessments of code quality may differ from benchmark results.

DeepSeek and Moonshot Accused of Secretly Routing Over 300, 000 Users to Claude. Search Engine Opened to Developers.

Alongside the new model, DeepSeek is opening its internal web search technology to outside developers for the first time.

Meanwhile, the new Search and Fetch APIs employ DeepSeek's own search engine. It the business notes already handles hundreds of millions of queries per day worldwide. Each endpoint costs $0.75 per 1, 000 calls.

DeepSeek API users can access the tools for free through server-side integrations in DeepSeek Harness, Claude Code and Codex until October 31, 2026, at 4: 00 PM UTC. China Could Restrict Open-Weight Models.

Notably, the release comes as the Chinese administration reportedly considers tighter controls on publishing open-weight AI models and other key AI technologies.

Under the documented proposal, whether a model ranks among the 10 most capable systems could be one factor in determining whether further restrictions apply.

That could complicate deeper reliance on Chinese open-weight models if tighter export or publication rules are introduced. Stay Connected with ProPakistani.

Obtain the latest tech news, telecom insights, and product launches wherever you prefer. Follow on Google Discover. Follow on Google News Join WhatsApp. See more ProPakistani stories in Google Search and Top Stories.

Technology and Automotive Specialist covering the latest cars, smartphones, AI breakthroughs, and.

In short, deepSeek V4.1 Flash Gets New Model With Big Coding Gains, Cheaper Agent is the central thread here, and readers can expect follow-up reporting as the picture becomes clearer.