Noometry

DeepSeek, open weights

DeepSeek V4.1 Flash

DeepSeek V4.1 Flash by DeepSeek ranks 38th of 354 ranked models on the Noometry Index as of October 2026, with a score of 52.8. Its strongest category is math, where it ranks 25th. API pricing starts at $0.15 per million input tokens and $0.60 per million output tokens, with a 1M-token context window.

Last verified

Noometry rank
#38 of 354
Index score
52.8
Evidence
Confirmed 37 results
Provider
DeepSeek
Released
September 9, 2026
Weights
Open weights
Reasoning
Yes
Context window
1M
Max output
393K
Input price
$0.15 / M
Output price
$0.60 / M
Blended price
$0.26 / M
Output speed
Not measured
Value
#42 of 219
Knowledge cutoff
May 2025
Input
text, image
Hugging Face
deepseek-ai/DeepSeek-V4.1-Flash

Category scores

Each category score combines every public result we have in that category.

  1. Coding 52.9
  2. Agentic & Tool Use 31.2
  3. Reasoning 50.2
  4. Math 66.7
  5. Knowledge 57.9
  6. Multimodal 39.1
  7. Multilingual 55.0
  8. Instruction Following 77.3
  9. Long Context 45.2
  10. Writing & Preference 65.4
CategoryScoreRankResults
Coding52.9#323
Agentic & Tool Use31.2#692
Reasoning50.2#367
Math66.7#255
Knowledge57.9#382
Multimodal39.1#612
Multilingual55.0#351
Instruction Following77.3#261
Long Context45.2#471
Writing & Preference65.4#484

Strengths and weaknesses

Categories where DeepSeek V4.1 Flash places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Math66.7+30.2#25 of 327, top 8%
Instruction Following77.3+6.1#26 of 305, top 9%
Coding52.9+14.2#32 of 340, top 10%

Weakest categories

CategoryScorevs medianRank
Multimodal39.1+0.5#61 of 128, top 48%
Agentic & Tool Use31.2+0.8#69 of 154, top 45%
Long Context45.2+4.2#47 of 296, top 16%

Closest competitors

The models ranked just above and below DeepSeek V4.1 Flash. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
GPT-5.2#3454.1$4.8115Compare
DeepSeek V4 Flash#3553.6$0.266Compare
GPT-6 Luna#3653.3$0.20—Compare
Grok 4.7#3753.1$3—Compare
GPT-5.2 Pro#3952.3$57.75—Compare
Gemini 3 Flash Preview#4052.3$1.13—Compare
GLM-5.3-Flash#4151.8$0.24—Compare
Qwen3.7 Max#4251.5$3.75—Compare

Sponsored placements are available on pages like this one. Advertise on Noometry

Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

Coding

BenchmarkScorePositionSettingSourceDate
LMArena WebDev1619#18 of 113, top 16%LMArena2026-10-08
SciCode51.9%#36 of 121, top 30%maxEpoch AI
SciCode35.5%noneEpoch AI
LMArena Coding1506#14 of 294, top 5%LMArena2026-10-08
ALE-Bench1,092#39 of 105, top 38%maxEpoch AI

Agentic & Tool Use

BenchmarkScorePositionSettingSourceDate
APEX-Agents39.5%#37 of 49, top 76%Epoch AI
GDP.pdf19.8%#20 of 36, top 56%maxEpoch AI

Reasoning

BenchmarkScorePositionSettingSourceDate
NYT Connections (extended)89.6%#19 of 91, top 21%high reasoningLech Mazur benchmarks
CritPt14.3%#38 of 134, top 29%maxEpoch AI
CritPt0.3%noneEpoch AI
LMArena Hard Prompts1483#29 of 297, top 10%LMArena2026-10-08
Mystery Game Puzzles43%#12 of 74, top 17%maxEpoch AI2026-10-08
DTBench89.9%#43 of 151, top 29%Epoch AI
LMCA47%#28 of 125, top 23%Epoch AI
Surface Evolver Bench46.3%#17 of 25, top 68%highEpoch AI
Epoch Capabilities Index154.9#31 of 213, top 15%Epoch AI2026-09-09

Math

BenchmarkScorePositionSettingSourceDate
FrontierMath (Tiers 1-3)67.4%#28 of 81, top 35%maxEpoch AI2026-10-08
FrontierMath Tier 426.8%#33 of 63, top 53%maxEpoch AI2026-10-08
OTIS Mock AIME 2024-202598.3%#21 of 173, top 13%maxEpoch AI2026-10-08
ProofBench54%#26 of 77, top 34%Epoch AI
LMArena Math1477#30 of 285, top 11%LMArena2026-10-08

Knowledge

BenchmarkScorePositionSettingSourceDate
GPQA Diamond89.8%#40 of 186, top 22%maxEpoch AI2026-10-09
LMArena Expert1506#20 of 273, top 8%LMArena2026-10-08

Multimodal

BenchmarkScorePositionSettingSourceDate
LMArena Vision1277#32 of 122, top 27%LMArena2026-10-09
Furniture Assembly34.2%#19 of 31, top 62%maxEpoch AI2026-10-08

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1448#35 of 297, top 12%LMArena2026-10-08
LMArena Chinese1497#47 of 285, top 17%LMArena2026-10-08
LMArena French1452#66 of 223, top 30%LMArena2026-10-08
LMArena German1484#18 of 231, top 8%LMArena2026-10-08
LMArena Japanese1412#48 of 211, top 23%LMArena2026-10-08
LMArena Korean1452#16 of 213, top 8%LMArena2026-10-08
LMArena Russian1471#26 of 283, top 10%LMArena2026-10-08
LMArena Spanish1459#36 of 226, top 16%LMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1474#23 of 298, top 8%LMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
LMArena Longer Query1475#29 of 291, top 10%LMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1462#32 of 297, top 11%LMArena2026-10-08
LMArena Creative Writing1435#44 of 295, top 15%LMArena2026-10-08
EQ-Bench Creative Writing1540#51 of 115, top 45%EQ-Bench
LMArena Multi-Turn1457#41 of 295, top 14%LMArena2026-10-08

API pricing by provider

RouteInput $/MOutput $/MCached input $/MChecked
deepinfra$0.20$0.60$0.0062026-10-10
deepseek$0.15$0.60$0.0032026-10-10
fireworks$0.30$1.20$0.0062026-10-10
openrouter$0.30$1.20$0.0062026-10-10
together$0.30$1.20$0.0062026-10-10

All DeepSeek API prices → Estimate your cost →

Compare DeepSeek V4.1 Flash

Other DeepSeek models

Frequently asked questions

How good is DeepSeek V4.1 Flash?

DeepSeek V4.1 Flash by DeepSeek ranks 38th of 354 ranked models on the Noometry Index as of October 2026, with a score of 52.8. Its strongest category is math, where it ranks 25th. API pricing starts at $0.15 per million input tokens and $0.60 per million output tokens, with a 1M-token context window.

How much does DeepSeek V4.1 Flash cost?

DeepSeek V4.1 Flash costs $0.15 per million input tokens and $0.60 per million output tokens on DeepSeek's own API, with cached input at $0.003.

What is DeepSeek V4.1 Flash's context window?

DeepSeek V4.1 Flash accepts up to 1M tokens of input and can write up to 393K tokens in one response.

Is DeepSeek V4.1 Flash open source?

Yes. DeepSeek V4.1 Flash's weights are downloadable from Hugging Face (deepseek-ai/DeepSeek-V4.1-Flash); check the license for commercial terms.

What are DeepSeek V4.1 Flash's strengths and weaknesses?

Relative to other ranked models, DeepSeek V4.1 Flash places best in math, instruction following, coding and lowest in multimodal, agentic & tool use, long context.

What is DeepSeek V4.1 Flash best at?

Its best category is math, where it ranks 25th on Noometry.