[whatmodel.sh](../../README.md) / DeepSeek V4.1 Flash 1017 # DeepSeek V4.1 Flash 1017 DeepSeek · released 2026-10 · mit **score 76.9** ±0.7 · #20 of 732 · from 17 benchmark results across 3 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks DeepSeek V4.1 Flash 1017 among every model on that task or test, worst to best. 
task         vs the others         rank  counts  tests
agent tasks  ▁▁▃▆█▄▅▅▄●▂▁    #18 of 313     23%      5
coding       ▁▁▁▃▅▇█▇▅▄●▁   #15 of 469*     10%      2
math         ▁▁▃▅█▆▄█▄●▁▁    #25 of 431     41%      3
agent tasks · #18 of 313 · 23% of its score Beats [Muse Wonder](muse-wonder.md), [GPT-5.5](gpt-5-5.md), [MiMo-V2.6-Pro](mimo-v2-6-pro.md) here, though they rank higher overall. Trails [Claude Opus 4.8](claude-opus-4-8.md) here, though they rank lower overall. 
test                                    result  vs the others in its run  difficulty            counts  why
Arena Agent (October 2026 aggrega…  0.0267 IPS  ▁▁▁▃▃▂●▁▂▂▁▁ beats 40/59  avg -0.379 IPS           13%  sharp
Terminal-Bench 4.0 (maker)               42.3%  ▁·▁··▁▆▆▅●█▃ beats 22/33  very hard · avg 0.7%    4.7%  sharp · at its level · small run
Chartography with tools (maker)          83.0%  ▂········▄●█ beats 3/10   very hard · avg 9.2%    3.5%  sharp · small run
DeepSWE v1.1 (resolved) (maker)          77.0%  ▂····▂▁▁▆██● beats 30/31  hard · avg 27%          1.2%  at its level · small run
Agents' Last Exam (metric/harness…       35.4%  ▂·▂·█▄●·▂▂▄▄ beats 8/14   very hard · avg 7.9%    1.0%  at its level · small run
coding · #15 of 469 · 10% of its score Beats [GPT-5.6 Sol](gpt-5-6-sol.md), [Claude Fable 5](claude-fable-5.md), [Muse Wonder](muse-wonder.md) here, though they rank higher overall. Trails [Qwen3.8 Max (0902)](qwen3-8-max-0902.md), [Qwen3.8 Max (2.4T-A95B)](qwen3-8-2-4t-a95b.md) here, though they rank lower overall. 
test                             result  vs the others in its run    difficulty      counts  why
Arena WebDev (Code Arena)        1665.0  ▁▁▃▂▅█▃▆▅●▁▂ beats 115/129  avg 981           9.1%  sharp
DeepSWE v1.1 (resolved) (maker)   77.0%  ▂····▂▁▁▆██● beats 30/31    hard · avg 27%    1.2%  at its level · small run
math · #25 of 431 · 41% of its score Trails [GPT-5.6 Luna](gpt-5-6-luna.md), [GPT-6 Luna](gpt-6-luna.md), [Claude Haiku 5.5](claude-haiku-5-5.md) here, though they rank lower overall. 
test                       result  vs the others in its run    difficulty            counts  why
FrontierMath tier 4 v2      46.3%  ▁▂▁▂█●▂▃▂▁▁▁ beats 45/67    very hard · avg 0.1%     23%  sharp · at its level
FrontierMath tiers 1–3 v2   71.2%  ▂··▁▁▃▄▄█●▄▃ beats 61/86    very hard · avg 1.2%     18%  sharp
OTIS mock AIME 2024–2025    99.4%  ▃▃▄▃▃▃▇▇█▃▂● beats 184/197  very hard · avg 7.2%    0.4%  sharp · maxed out · noisy
other tests (8) 
test                            result  vs the others in its run    difficulty            counts  why
Mystery Game Puzzles (Epoch)     49.0%  █··▂▃▇▄█●▄▂▁ beats 65/77    very hard · avg 9.4%     10%  sharp · at its level · noisy
Chess Puzzles (Epoch)            46.0%  █▁▁▁▁▂▃▃▃●▁▁ beats 125/137  very hard · avg 5.8%    5.4%  sharp · at its level · noisy
GPQA Diamond                     93.2%  ▁▁▁▁▂▄▆▄▅▇█● beats 195/211  hard · avg 39%          2.5%  sharp · maxed out
Humanity's Last Exam (maker)     42.6%  ▁▁▁▄▆▆▇██●▁▃ beats 98/111   very hard · avg 4.6%    2.0%  at its level · small run
SimpleQA Verified                42.5%  ▁▁▃▂▃▃▆●█▄▂▅ beats 39/81    very hard · avg 8.0%    2.0%  at its level
AutomationBench v1.0.6 (maker)   60.5%  ▃···▃·▃▃··█● beats 8/8      very hard · avg 9.9%    1.7%  at its level · small run
NL2Repo (maker)                  68.2%  ▂▂▂·▃▃▅█▃▂▅● beats 22/22    hard · avg 15%          1.5%  at its level · small run
BabyVision with tools (maker)    92.5%  █······█··█● beats 3/3      medium · avg 49%        1.2%  small run
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
Epoch AI                                        7   61% ▇▇▇▇▇▇▇
LMArena leaderboard dataset                     2   22% ▇▇▇
DeepSeek V4.1 Flash 1017 model card (su…        8   17% ▇▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 76.9
Licences: Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · DeepSeek V4.1 Flash 1017 model card ([MIT](../../licenses/deepseek-1017-card.txt)) ## How to run it ### Through an API api (DeepSeek) · $0.26/M blended · $0.15 in / $0.6 out per 1M tokens · [api-docs.deepseek.com](https://api-docs.deepseek.com/quick_start/pricing) [price source](https://api-docs.deepseek.com/quick_start/pricing): USD per 1M tokens, off-peak uncached input/output. Peak rates are $0.30/$1.20, 01:00–04:00 and 06:00–10:00 UTC weekdays excluding Chinese public holidays. Cached input excluded. API deepseek-flash serves the 1017 checkpoint; old Flash aliases also route here. Observed 2026-10-21. · [context source](https://api-docs.deepseek.com/quick_start/pricing): 1M tokens per official API pricing/specification, observed 2026-10-21. Model repository and documentation: [huggingface.co/deepseek-ai](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017) [interactive version](https://whatmodel.sh/model/deepseek-v4-1-flash-1017) · [JSON](https://whatmodel.sh/api?format=json&all=1)