What is the IQ of minimax-m3?

minimax-m3 has an estimated IQ of 114, which ranks #64 out of 176 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 minimax-m3 best at?

Among the dimensions with published benchmark results, minimax-m3 ranks highest in Reliability (IQ 123, #29 of 177) and Mathematical Reasoning (IQ 120, #63 of 176). Its lowest-ranked measured dimension is Abstract Reasoning (IQ 86, #130 of 201).

What is the EQ of minimax-m3?

minimax-m3 has an estimated EQ of 104. EQ measures emotional reasoning — how well a model reads and responds to emotional context — and is scored separately from IQ.

Which benchmarks has minimax-m3 been tested on?

minimax-m3 has published results on 23 of the 39 benchmarks behind AI IQ's Composite IQ, including IFBench, GPQA Diamond, AA Long Context Reasoning v1.1, BullshitBench v2 and Humanity's Last Exam.

How much does minimax-m3 cost?

minimax-m3 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 $3.43 per million input-plus-output tokens.

Is minimax-m3 open source?

Yes — minimax-m3 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.

Who developed minimax-m3?

minimax-m3 was developed by MiniMax, based in China.

When was minimax-m3 released?

minimax-m3 was publicly released on June 1, 2026.

How large is the context window of minimax-m3?

minimax-m3 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.

How fast is minimax-m3?

In independent measurements, minimax-m3 generates about 109 tokens per second, with a typical response time of about 24 seconds.

MiniMaxReleased Jun 1, 2026

Open weights

minimax-m3

Strongest in Reliability (#29 of 177) and Mathematical Reasoning (#63 of 176); weakest in Abstract Reasoning (#130 of 201). All seven dimensions are measured.

Composite IQ ?

114

#64 of 176 ranked models · ahead of 64%

MiniMax's #2 model, behind MiniMax M3.1 Flash (120)

Effective cost

$3.43/1M

$0.30 in · $1.20 out

Speed

109tok/s

24s typical response

Context

1Mtokens

1,000,000 tokens in

Evidence

7of 7 dims

23 of 39 benchmarks

Compare with gpt-6.1-solAPI pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

Measured95% 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

Abstract ReasoningAbstract

86measured

#130 / 201

2/5

-40.5

Mathematical ReasoningMath

120measured

#63 / 176

1/5

-27

Academic ReasoningAcademic

125measured

#76 / 203

5/6

-10.5

Programmatic ReasoningCoding

118measured

#65 / 177

2/6

-31.5

Computer UseComputer Use

115measured

#70 / 176

3/5

-28

ReliabilityReliability

123measured

#29 / 177

4/6

-11

Knowledge WorkKnowledge Work

113measured

#70 / 176

6/6

-18.5

minimax-m395% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

23 of 39 Composite IQ benchmarks have a published result for minimax-m3. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 86 · measuredBest: #63 of 76 on Chess Puzzles2/5

BenchmarkScorePercentileRankLeader · source

Chess Puzzles

Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 14

14.0%

#63 / 76

Leader gpt-6-astra · 72.0%
Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 14

Mystery Game Puzzles

Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

7.0%

#52 / 55

Leader gpt-6-astra · 84.0%
Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

Not yet reported

ARC-AGI-3ARC-AGI-2ARC-AGI-1

Mathematical Reasoning IQ 120 · measuredBest: #41 of 48 on ProofBench 1.11/5

BenchmarkScorePercentileRankLeader · source

ProofBench 1.1

Full source leaderboard reconciliation v1.1 · Oct 27

18.0%

#41 / 48

Leader fable-5.1 · 100%
Full source leaderboard reconciliation v1.1 · Oct 27

Not yet reported

FrontierMath Tier 4FrontierMath Tier 1-3MathArenaAIME

Academic Reasoning IQ 125 · measuredBest: #18 of 147 on GPQA Diamond5/6

BenchmarkScorePercentileRankLeader · source

GPQA Diamond

Artificial Analysis v4.3.2 · Oct 27

92.9%

#18 / 147

Leader gpt-6-astra · 96.1%
Artificial Analysis v4.3.2 · Oct 27

Humanity's Last Exam

Artificial Analysis v4.3.2 · Oct 27

39.0%

#55 / 167

Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 27

MMMU-Pro

Artificial Analysis v4.3.2 · Oct 27

78.6%

#36 / 89

Leader fable-5.1 · 90.6%
Artificial Analysis v4.3.2 · Oct 27

SciCode

Artificial Analysis v4.3.2 · Oct 27

47.1%

#80 / 167

Leader opus-5.5 · 66.9%
Artificial Analysis v4.3.2 · Oct 27

CritPt

Artificial Analysis v4.3.2 · Oct 27

3.7%

#89 / 159

Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

MMLU-Pro

Programmatic Reasoning IQ 118 · measuredBest: #54 of 104 on Terminal-Bench 4.02/6

BenchmarkScorePercentileRankLeader · source

Terminal-Bench 4.0

Artificial Analysis v4.3.2 · Oct 27

2.0%

#54 / 104

Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v4.3.2 · Oct 27

SWE-rebench

Full source leaderboard reconciliation · Oct 23

47.2%

#16 / 25

Leader fable-5 · 64.5%
Full source leaderboard reconciliation · Oct 23

Not yet reported

LiveCodeBenchIOI V2ProgramBenchMirrorCode

Computer Use IQ 115 · measuredBest: #15 of 43 on BrowseComp3/5

BenchmarkScorePercentileRankLeader · source

BrowseComp

Full source leaderboard reconciliation · Oct 23

83.5%

#15 / 43

Leader Kimi K3.1 Pro · 92.1%
Full source leaderboard reconciliation · Oct 23

OSWorld-Verified

Full source leaderboard reconciliation · Oct 23

70.1%

#18 / 24

Leader fable-5 · 85.0%
Full source leaderboard reconciliation · Oct 23

Arena.ai Agent Arena

Arena.ai Agent Arena · Oct 26

-10.2

#62 / 70

Leader opus-5.5 · 20.9
Arena.ai Agent Arena · Oct 26

Not yet reported

MCP AtlasAgents' Last Exam

Reliability IQ 123 · measuredBest: #1 of 96 on IFBench4/6

BenchmarkScorePercentileRankLeader · source

IFBench

Artificial Analysis v4.3.2 · Oct 27

82.9%

#1 / 96

Leader on this benchmark
Artificial Analysis v4.3.2 · Oct 27

AA Long Context Reasoning v1.1

Artificial Analysis v4.3.2 · Oct 27

83.0%

#21 / 154

Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 27

BullshitBench v2

Full source leaderboard reconciliation · Oct 23

64.0%

#16 / 83

Leader opus-4.8 · 94.0%
Full source leaderboard reconciliation · Oct 23

AA Omniscience

Artificial Analysis v4.3.2 · Oct 27

1.4

#64 / 158

Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

SimpleQA VerifiedFACTS Grounding

Knowledge Work IQ 113 · measuredBest: #51 of 124 on GDPval-AA v2.1 normalized score6/6

BenchmarkScorePercentileRankLeader · source

GDPval-AA v2.1 normalized score

Artificial Analysis v4.3.2 · Oct 23

37.1%

#51 / 124

Leader opus-5.5 · 68.3%
Artificial Analysis v4.3.2 · Oct 23

AA-Briefcase v1.1 Elo

Artificial Analysis v4.3.2 · Oct 27

1,089

#50 / 111

Leader sonnet-5.5 · 1,812
Artificial Analysis v4.3.2 · Oct 27

AutomationBench-AA

Artificial Analysis v4.3.2 · Oct 27

21.3%

#57 / 106

Leader Gemini 4 Argon · 77.5%
Artificial Analysis v4.3.2 · Oct 27

GDP.pdf

Artificial Analysis v4.3.2 · Oct 27

9.8%

#64 / 106

Leader Muse Wonder Full · 32.6%
Artificial Analysis v4.3.2 · Oct 27

EnterpriseOps-Gym-AA

Full source leaderboard reconciliation v1117-task-strict-2026-09-29 · Oct 19

32.1%

#30 / 31

Leader fable-5 · 51.1%
Full source leaderboard reconciliation v1117-task-strict-2026-09-29 · Oct 19

AA-AnalystAgent

Full source leaderboard reconciliation vpass^5-2026-09-29 · Oct 19

10.0%

#34 / 35

Leader gemini-3.7-flash · 60.0%
Full source leaderboard reconciliation vpass^5-2026-09-29 · Oct 19

Release & availability

Release details on record for this model.

Released: Jun 1, 2026
Developer: MiniMax
Weights: Open weightsDownloadable and self-hostable
Context: 1,000,000 tokens
API price: $0.30 in · $1.20 outper 1M tokens · full pricing

About this score

How much of minimax-m3's IQ is measured, and how the rest is estimated.

23/ 39benchmark results

7/ 7dimensions measured

0estimated

2/5Abstract

1/5Math

5/6Academic

2/6Coding

3/5Computer Use

4/6Reliability

6/6Knowledge 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.

How gaps are handled

Every dimension has at least one published result.

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.

Read the scoring methodology →

Compare minimax-m3

Side-by-side pages for the models it is most often weighed against.

#1 overall

gpt-6.1-sol

IQ141 27 higherEffective cost$6.30 1.8× pricier

MiniMax #1

MiniMax M3.1 Flash

IQ120 6 higherEffective cost$1.42 2.4× cheaper

IQ113 1 lowerEffective cost$10.62 3.1× pricier

Similar price

kimi-k2.7-code

IQ111 3 lowerEffective cost$3.57 about the same

#2 overall

opus-5.5

IQ141 27 higherEffective cost$35.73 10.4× pricier

minimax-m3Comparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about minimax-m3, 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 minimax-m3?: minimax-m3 has an estimated IQ of 114, which ranks #64 out of 176 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 minimax-m3 best at?: Among the dimensions with published benchmark results, minimax-m3 ranks highest in Reliability (IQ 123, #29 of 177) and Mathematical Reasoning (IQ 120, #63 of 176). Its lowest-ranked measured dimension is Abstract Reasoning (IQ 86, #130 of 201).
What is the EQ of minimax-m3?: minimax-m3 has an estimated EQ of 104. EQ measures emotional reasoning — how well a model reads and responds to emotional context — and is scored separately from IQ.
Which benchmarks has minimax-m3 been tested on?: minimax-m3 has published results on 23 of the 39 benchmarks behind AI IQ's Composite IQ, including IFBench, GPQA Diamond, AA Long Context Reasoning v1.1, BullshitBench v2 and Humanity's Last Exam.
How much does minimax-m3 cost?: minimax-m3 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 $3.43 per million input-plus-output tokens.
Is minimax-m3 open source?: Yes — minimax-m3 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
Who developed minimax-m3?: minimax-m3 was developed by MiniMax, based in China.
When was minimax-m3 released?: minimax-m3 was publicly released on June 1, 2026.
How large is the context window of minimax-m3?: minimax-m3 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
How fast is minimax-m3?: In independent measurements, minimax-m3 generates about 109 tokens per second, with a typical response time of about 24 seconds.