Compare Models

DeepSeek V4.1 Flash 1017

DeepSeek V4.1 Flash 1017 is an open-weight model from DeepSeek, released October 17, 2026. It ranks #14 of 50 models on the Vals Index with 56.58%. Its best result is #5 of 71 on Tax Agent Bench.

Release Date: Oct 17, 2026

Developer DeepSeek  🇨🇳
Context Window 1M
Max Output Tokens 384k
Token Costs (in/out) $0.30/1.20
Weights Open
Input Modalities

Accuracy

56.58 % ± 1.27

Cost / Test (Vals Index)

$ 0.637

Latency

36 min 33 s

Vals Index
BenchmarksAccuracyRankings
Vals Index

56.58%

±1.27
14/50
Code Migration

44.21%

±4.46
25/79
EMB

61.80%

±3.19
32/76
Finance Agent (v2)

56.96%

±2.05
17/80
Legal Research Bench

45.19%

±3.46
17/80
Tax Agent Bench

44.35%

±3.57
5/71
Vibe Code Bench v1.1

87.73%

±2.75
11/114
Harvey's Legal Agent Benchmark

9.17%

±2.51
22/81
Terminal-Bench 4.0

31.82%

±2.62
17/50
Proprietary BenchmarksAcademic BenchmarksIndustry Partners
Vals
Default Provider : DeepSeek
Temperature: 1
Top P: 1
Top K: Default
Max Output Tokens: 384,000
Reasoning Effort: max

Compared with

DeepSeek V4.1 Flash 1017 vs GPT-6 Sol 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs GPT-5.6 Sol 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs Muse Spark 1.3 Max 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs MiMo V2.6 Pro 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs Claude Opus 4.8 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs Grok 4.7 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs Gemini 3.8 Flash 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs Claude Haiku 5.5 9 shared benchmarks DeepSeek V4.1 Flash 1017 vs GPT-6.1 Sol 9 shared benchmarks