DeepSeek V4.1 Flash (high): benchmark scores, price and ranking
DeepSeek V4.1 Flash (high) from DeepSeek ranks #317 of 743 LLMs on the LLMs Tiger aggregate benchmark score (48.9), based on 3 benchmark results captured up to 2026-10-13.
- Organisation
- DeepSeek
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2026-09-09 (source: epoch)
- Cheapest API price
- $0.010 in / $1.20 out per million tokens (via coralbricks, observed 2026-10-13T20:47Z)
- Aggregate rank
- #317 of 743 (score 48.9)
Across the 2 benchmarks where at least ten models are scored, DeepSeek V4.1 Flash (high) is in the top 10% on 0 and in the top half on 1.
Its cheapest listed input price, $0.010 per million tokens, is lower than that of 99% of the 427 models with a listed price.
DeepSeek V4.1 Flash (high) benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| ARC-AGI-2 | 67.5% | #21 of 87 | ARC Prize official leaderboard | 2026-10-13 |
| Surface Evolver | 46.2% | #7 of 13 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WeirdML v3 | 5.9% | #3 of 3 | Epoch AI Benchmarking Hub | 2026-10-13 |
About DeepSeek V4.1 Flash (high)
DeepSeek V4.1 Flash (high) is an open-weight model that can run on your own hardware from DeepSeek. It was released on 2026-09-09, according to Epoch AI. The cheapest API listing we track charges $0.010 per million input tokens and $1.20 per million output tokens, through coralbricks. We hold 3 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where DeepSeek V4.1 Flash (high) stands
On our aggregate score, DeepSeek V4.1 Flash (high) is #317 of 743 ranked models, around the middle of the table. Its score is 48.9; the scale is centred on 50, so it sits 1.1 points below the typical model. That beats 57% of the ranked field. Among the 33 ranked DeepSeek models it is #15. The next one up is deepseek-v4-pro-high-20260813 at 48.9. The next one down is DeepSeek-R1 (May 2025) at 48.2.
DeepSeek V4.1 Flash (high), benchmark by benchmark
ARC-AGI-2 puts DeepSeek V4.1 Flash (high) at 67.5%, #21 of 87. That beats 76% of the models tested, but 27.5 points short of GPT-6 Astra (max) (95.0%). The results either side are Claude Opus 4.6 (120K, Max) (68.8%) and Grok 4.6 (xhigh) (67.1%). It covers abstract reasoning, ARC Prize public eval set. The board median is 22.8%, so DeepSeek V4.1 Flash (high) is 44.7 points above it. Source: ARC Prize official leaderboard, captured 2026-10-13.
DeepSeek V4.1 Flash (high) lands below the middle on Surface Evolver: 46.2%, #7 of 13, with 46% of the field scoring higher. Kimi K3 (max) leads at 93.0%; the closest result above is Kimi K2.7 Code at 48.8%. It covers geometry optimisation with a solver. The median score on this board is also 46.2%. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
The full table above also lists the other 1 result, including boards with fewer than ten models, where a rank says little.
Strengths and weaknesses
DeepSeek V4.1 Flash (high) has too few results on boards with ten or more models to say where it is strong or weak. Treat any single score as a data point, not a verdict.
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so DeepSeek V4.1 Flash (high) costs less than the typical listing. We found no cheaper model with a similar or better aggregate score, so DeepSeek V4.1 Flash (high) is the lowest-priced option at its level. Similar scores cost more on GLM-5.3-Flash (max) ($0.026) and Qwen3.8-27B (XHigh) ($0.040).
Release history
The DeepSeek model released before it was deepseek-v4-pro-high-20260813 on 2026-08-13, which scores 48.9. The next one was deepseek-v4.1-flash-max on 2026-09-09, at 52.2.
Reading these numbers
The 3 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (2), ARC Prize official leaderboard (1). 1 of them come from an official leaderboard and 2 from datasets or third-party runs. The leaderboard also lists 4 other reasoning-effort settings of this model, which redirect here. This page shows the highest. With 3 results the aggregate rank can move a lot when one new score arrives.
DeepSeek V4.1 Flash (high) is also listed at 4 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.
Models ranked nearby
- #314 Inkling (xhigh) (48.9)
- #315 grok-4.3 (high) (48.9)
- #316 claude-opus-4-5-20251101-high-32k (48.9)
- #318 Gemini 3.6 Flash (High) (48.9)
- #319 Claude Opus 4 (48.9)
- #320 Gemini 2.0 Flash (Experimental) (48.8)
More from DeepSeek
- DeepSeek V4 Pro 0813 (max) (#23, 54.3)
- DeepSeek V4 Flash 0731 (max) (#42, 53.1)
- DeepSeek-V3.2-Exp (high) (#67, 52.4)
- deepseek-v4.1-flash-max (#75, 52.2)
- DeepSeek-V2.5-1210 (#101, 51.7)
- DeepSeek v4 Pro (max) (#108, 51.6)
Compare it on the live leaderboard → · How the aggregate score works