Skip to content
DigIntoAI/Model intelligence
Model intelligence feed

Model intelligence.

Follow the changes. Understand what comes next.

Editor’s pickDeepSeek
DeepSeek/DeepSeek V4.1 Flash 1017;ARC-AGI-3 Standard/Provider Adapter harness

Why Does the Same DeepSeek Score So Differently Across ARC-AGI-3 Harnesses?

On the ARC Prize official results page for DeepSeek V4.1 Flash 1017 (page label Oct 17, 2026 · 5 harness configurations), the same model at the same max reasoning setting scores 0.2% on ARC-AGI-3 under Standard and 5.8% under Provider Adapter. The gap is the evaluation harness—kept notes versus opaque state plus compaction—not a new weight drop. Same-page Semi-Private peaks: v1 high 94.5%, v2 max 78.1%. Figures are ARC Prize self-reports; we haven't retested.

Latest stories03

Newest published
01

Why Does the Same DeepSeek Score So Differently Across ARC-AGI-3 Harnesses?

On the ARC Prize official results page for DeepSeek V4.1 Flash 1017 (page label Oct 17, 2026 · 5 harness configurations), the same model at the same max reasoning setting scores 0.2% on ARC-AGI-3 under Standard and 5.8% under Provider Adapter. The gap is the evaluation harness—kept notes versus opaque state plus compaction—not a new weight drop. Same-page Semi-Private peaks: v1 high 94.5%, v2 max 78.1%. Figures are ARC Prize self-reports; we haven't retested.

DeepSeek V4.1 Flash 1017;ARC-AGI-3 Standard/Provider Adapter harnessRead story
02

Same Open Model, Different API: Why Speed and Cost Both Change

Vals AI’s Inference Provider Index (page Updated 10/20/2026; Updates dated 10/22/2026) runs the same open-weight models—DeepSeek V4.1 Flash, Kimi K3, and MiniMax M3—on first-party APIs and hosts such as Together AI, Fireworks, and Baseten. Across 165 agent tasks, overall accuracy stayed within error of the first-party API. Cost per task rose by as much as about 2.3× by host, and for all three models the creator’s API was neither the cheapest nor the fastest. Figures below are from Vals AI, for reference only.

Inference Provider IndexRead story
03

Is DeepSeek Banned in the US? What the New DHS Order Covers

DeepSeek is not banned across the United States. A DHS order published in the Federal Register on October 20, 2026 applies to federal civilian executive branch agencies: it excludes DeepSeek products from their procurements and requires removal of DeepSeek products, and third-party products built on DeepSeek implementations, from their IT systems. National security systems are outside its scope; agencies can request exceptions. Texas, New York and Australia have separate rules for government devices; Italy's privacy regulator restricted the chatbot's processing of Italian users' data in 2025.

DeepSeekRead story