DeepSeekReleased Oct 17, 2026
NewOpen weights
DeepSeek V4.1 Flash 1017
Composite IQ ?
125
Effective cost
$1.87/1M
$0.30 in · $1.20 out
Speed
Not measured
No speed data yet
Context
1Mtokens
1,000,000 tokens in
Evidence
0of 7 dims
0 of 39 benchmarks
IQ profile
Seven dimensions, averaged equally into Composite IQ
MeasuredEstimated95% range
#1 in dimensionTop-10 median
IQ dimensions
Ranks are among models scored in each dimension. Ranges are model-based estimates, not calibrated confidence intervals.
DimensionIQPosition · 70–160RankEvidencevs Top 10
DeepSeek V4.1 Flash 101795% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human
Benchmark results
0 of 39 Composite IQ benchmarks have a published result for DeepSeek V4.1 Flash 1017. Benchmarks without a result stay listed so gaps are visible.
Abstract Reasoning IQ 112 · estimatedNo published results0/5
No published results yet for Abstract Reasoning. The 112 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
ARC-AGI-3ARC-AGI-2ARC-AGI-1Chess PuzzlesMystery Game Puzzles
Mathematical Reasoning IQ 130 · estimatedNo published results0/5
No published results yet for Mathematical Reasoning. The 130 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
FrontierMath Tier 4FrontierMath Tier 1-3ProofBench 1.1MathArenaAIME
Academic Reasoning IQ 130 · estimatedNo published results0/6
No published results yet for Academic Reasoning. The 130 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
Humanity's Last ExamGPQA DiamondCritPtSciCodeMMLU-ProMMMU-Pro
Programmatic Reasoning IQ 135 · estimatedNo published results0/6
No published results yet for Programmatic Reasoning. The 135 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
LiveCodeBenchIOI V2Terminal-Bench 4.0ProgramBenchSWE-rebenchMirrorCode
Computer Use IQ 126 · estimatedNo published results0/5
No published results yet for Computer Use. The 126 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
BrowseCompOSWorld-VerifiedMCP AtlasArena.ai Agent ArenaAgents' Last Exam
Reliability IQ 119 · estimatedNo published results0/6
No published results yet for Reliability. The 119 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
SimpleQA VerifiedaaOmniscienceBullshitBench v2IFBenchAA Long Context Reasoning v1.1FACTS Grounding
Knowledge Work IQ 125 · estimatedNo published results0/6
No published results yet for Knowledge Work. The 125 shown is estimated from DeepSeek V4.1 Flash 1017's other dimensions and will be replaced as results come in. How estimates work
AA-Briefcase v1.1GDPval-AA v2.1AutomationBench-AAGDP.pdfAA-AnalystAgentEnterpriseOps-Gym-AA
Release & availability
Facts from the publisher's announcement and documentation.
- Released
- Developer
- DeepSeek
- Access
- MIT checkpoint weights and the DeepSeek API as deepseek-flash; web and app also route to this update.
- Weights
- Open weightsDownloadable and self-hostable
- Inputs
- textimage
- Context
- 1,000,000 tokens
- API price
- $0.30 in · $1.20 outper 1M tokens · full pricing
What DeepSeek says
- DeepSeek released the dated V4.1 Flash 1017 checkpoint on October 17, 2026.
- The publisher reports improvements to reasoning, coding and visual tasks with reduced output length; API calling method and prices stay unchanged.
Limitations
- Stored prices use peak rates; off-peak rates are half the peak rates.
- Provider harness scores have not been verified as compatible canonical benchmark measurements.
- The older September checkpoint remains separately downloadable; its benchmark observations are preserved on that checkpoint.
Publisher statements, not AI IQ measurements.
About this score
How much of DeepSeek V4.1 Flash 1017's IQ is measured, and how the rest is estimated.
0/ 39benchmark results
0/ 7dimensions measured
7estimated
0/5Abstract
0/5Math
0/6Academic
0/6Coding
0/5Computer Use
0/6Reliability
0/6Knowledge Work
Bar height = share of benchmarks with resultsEstimatedNo score
Measured results are shown as published
Every score in the benchmark tables comes from its source unchanged. Composite IQ averages all seven dimensions equally.
Estimates fill the gaps
Abstract Reasoning, Mathematical Reasoning, Academic Reasoning, Programmatic Reasoning, Computer Use, Reliability and Knowledge Work have no published results yet, so their IQ is estimated from DeepSeek V4.1 Flash 1017's other dimensions. Missing benchmarks can also draw on its predecessor, Deepseek V4.1 Flash; a predecessor is an estimation input, not evidence that this model has the predecessor's measured results.
A missing result is not a zero
Benchmark versions and evaluation harnesses can differ between sources, so small gaps between models may not be meaningful.
Compare DeepSeek V4.1 Flash 1017
Side-by-side pages for the models it is most often weighed against.
#1 overall
gpt-6.1-sol
IQ141 16 higherEffective cost$6.30 3.4× pricier
Predecessor
DeepSeek V4.1 Flash
IQ126 1 higherEffective cost$1.87 about the same
Closest IQ
gpt-6-luna
IQ125 same IQEffective cost$0.771 2.4× cheaper
Most similar profile
mimo-v2.6-pro
IQ125 same IQEffective cost$1.52 1.2× cheaper
Similar price
grok-4.3
IQ112 13 lowerEffective cost$1.69 1.1× cheaper
#2 overall
opus-5.5
IQ141 16 higherEffective cost$35.73 19.1× pricier
DeepSeek V4.1 Flash 1017Comparison model
Frequently asked questions
Quick answers about DeepSeek V4.1 Flash 1017, drawn from the same data as the rest of this page.
How AI IQ scores models →Get this data from the API →- What is the IQ of DeepSeek V4.1 Flash 1017?
- DeepSeek V4.1 Flash 1017 has an estimated IQ of 125, which ranks #24 out of 169 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.
- What is DeepSeek V4.1 Flash 1017 best at?
- There isn't any published benchmark data to judge that yet. None of the seven dimensions AI IQ scores has a measured result for DeepSeek V4.1 Flash 1017, so its strengths and weaknesses can't be compared with other models.
- What is the EQ of DeepSeek V4.1 Flash 1017?
- DeepSeek V4.1 Flash 1017 doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for DeepSeek V4.1 Flash 1017. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
- Which benchmarks has DeepSeek V4.1 Flash 1017 been tested on?
- None yet. DeepSeek V4.1 Flash 1017 has no published results on the 39 benchmarks behind AI IQ's Composite IQ. AI IQ only uses published, source-backed results, so its profile fills in as evaluations of this model appear.
- How much does DeepSeek V4.1 Flash 1017 cost?
- DeepSeek V4.1 Flash 1017 costs $0.30 per million input tokens and $1.20 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $1.87 per million input-plus-output tokens.
- Is DeepSeek V4.1 Flash 1017 open source?
- Yes — DeepSeek V4.1 Flash 1017 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
- Who developed DeepSeek V4.1 Flash 1017?
- DeepSeek V4.1 Flash 1017 was developed by DeepSeek, based in China.
- When was DeepSeek V4.1 Flash 1017 released?
- DeepSeek V4.1 Flash 1017 was publicly released on October 17, 2026.
- How large is the context window of DeepSeek V4.1 Flash 1017?
- DeepSeek V4.1 Flash 1017 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
- How fast is DeepSeek V4.1 Flash 1017?
- AI IQ doesn't have independent speed measurements for DeepSeek V4.1 Flash 1017 yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.