What is the IQ of hy3?

hy3 has an estimated IQ of 99, which ranks #133 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.

Among the dimensions with published benchmark results, hy3 ranks highest in Computer Use (IQ 117, #62 of 176) and Academic Reasoning (IQ 121, #96 of 203). Its lowest-ranked measured dimension is Reliability (IQ 98, #154 of 177).

What is the EQ of hy3?

hy3 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 hy3. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.

Which benchmarks has hy3 been tested on?

hy3 has published results on 9 of the 39 benchmarks behind AI IQ's Composite IQ, including BrowseComp, GPQA Diamond, CritPt, Humanity's Last Exam and SciCode.

How much does hy3 cost?

hy3 costs $0.20 per million input tokens and $0.80 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $0.6543 per million input-plus-output tokens.

Yes — hy3 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.

hy3 was developed by Tencent, based in China.

When was hy3 released?

hy3 was publicly released on April 24, 2026.

How large is the context window of hy3?

hy3 supports a context window of 256,000 tokens — the amount of text it can consider in a single request.

In independent measurements, hy3 generates about 195 tokens per second, with a typical response time of about 4.3 seconds.

TencentReleased Apr 24, 2026

Open weights

hy3

Its best results are in Computer Use (#62 of 176) and Academic Reasoning (#96 of 203); weakest in Reliability (#154 of 177). Four of seven dimensions are estimated, because no results are published yet.

Composite IQ ?

#133 of 176 ranked models · ahead of 25%

China (Other)'s #11 model, behind Ling 3.1 Flash (120)

Effective cost

$0.654/1M

$0.20 in · $0.80 out

Speed

195tok/s

4s typical response

Context

256Ktokens

256,000 tokens in

Evidence

3of 7 dims

9 of 39 benchmarks

Compare with gpt-6.1-solAPI pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

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

Abstract ReasoningAbstract

78estimated

#167 / 201

0/5

-48.5

Mathematical ReasoningMath

94estimated

#138 / 176

0/5

-53

Academic ReasoningAcademic

121measured

#96 / 203

4/6

-14.5

Programmatic ReasoningCoding

92estimated

#162 / 177

0/6

-57.5

Computer UseComputer Use

117measured

#62 / 176

2/5

-26

ReliabilityReliability

98measured

#154 / 177

3/6

-36

Knowledge WorkKnowledge Work

94estimated

#142 / 176

0/6

-37.5

hy395% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

9 of 39 Composite IQ benchmarks have a published result for hy3. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 78 · estimatedNo published results0/5

No published results yet for Abstract Reasoning. The 78 shown is estimated from hy3'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 94 · estimatedNo published results0/5

No published results yet for Mathematical Reasoning. The 94 shown is estimated from hy3'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 121 · measuredBest: #69 of 147 on GPQA Diamond4/6

BenchmarkScorePercentileRankLeader · source

GPQA Diamond

Artificial Analysis model leaderboard · Sep 10

87.0%

#69 / 147

Leader gpt-6-astra · 96.1%
Artificial Analysis model leaderboard · Sep 10

CritPt

Artificial Analysis model leaderboard · Sep 10

5.0%

#81 / 159

Leader Muse Wonder Full · 32.9%
Artificial Analysis model leaderboard · Sep 10

Humanity's Last Exam

Artificial Analysis model leaderboard · Sep 10

26.0%

#102 / 167

Leader opus-5.5 · 61.4%
Artificial Analysis model leaderboard · Sep 10

SciCode

Artificial Analysis model leaderboard · Sep 10

41.0%

#113 / 167

Leader opus-5.5 · 66.9%
Artificial Analysis model leaderboard · Sep 10

Not yet reported

MMLU-ProMMMU-Pro

Programmatic Reasoning IQ 92 · estimatedNo published results0/6

No published results yet for Programmatic Reasoning. The 92 shown is estimated from hy3's other dimensions and will be replaced as results come in. How estimates work

LiveCodeBenchIOI V2Terminal-Bench 4.0ProgramBenchSWE-rebenchMirrorCode

Computer Use IQ 117 · measuredBest: #12 of 43 on BrowseComp2/5

BenchmarkScorePercentileRankLeader · source

BrowseComp

Full source leaderboard reconciliation · Oct 23

84.2%

#12 / 43

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

Arena.ai Agent Arena

Arena.ai Agent Arena · Oct 26

-7.3

#55 / 70

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

Not yet reported

OSWorld-VerifiedMCP AtlasAgents' Last Exam

Reliability IQ 98 · measuredBest: #65 of 96 on IFBench3/6

BenchmarkScorePercentileRankLeader · source

IFBench

Artificial Analysis model leaderboard · Aug 19

63.0%

#65 / 96

Leader minimax-m3 · 82.9%
Artificial Analysis model leaderboard · Aug 19

AA Omniscience

Artificial Analysis model leaderboard · Sep 10

-35.0

#130 / 158

Leader opus-5.5 · 46.4
Artificial Analysis model leaderboard · Sep 10

BullshitBench v2

Full source leaderboard reconciliation · Oct 23

8.0%

#78 / 83

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

Not yet reported

SimpleQA VerifiedAA Long Context Reasoning v1.1FACTS Grounding

Knowledge Work IQ 94 · estimatedNo published results0/6

No published results yet for Knowledge Work. The 94 shown is estimated from hy3'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

Release details on record for this model.

Released: Apr 24, 2026
Developer: Tencent
Weights: Open weightsDownloadable and self-hostable
Context: 256,000 tokens
API price: $0.20 in · $0.80 outper 1M tokens · full pricing

About this score

How much of hy3's IQ is measured, and how the rest is estimated.

9/ 39benchmark results

3/ 7dimensions measured

4estimated

0/5Abstract

0/5Math

4/6Academic

0/6Coding

2/5Computer Use

3/6Reliability

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, Programmatic Reasoning and Knowledge Work have no published results yet, so their IQ is estimated from hy3's other dimensions.

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 hy3

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

#1 overall

gpt-6.1-sol

IQ141 42 higherEffective cost$6.30 9.6× pricier

IQ99 same IQEffective cost$0.137 4.8× cheaper

IQ114 15 higherEffective cost$0.647 about the same

#2 overall

opus-5.5

IQ141 42 higherEffective cost$35.73 54.6× pricier

hy3Comparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about hy3, 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 hy3?: hy3 has an estimated IQ of 99, which ranks #133 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 hy3 best at?: Among the dimensions with published benchmark results, hy3 ranks highest in Computer Use (IQ 117, #62 of 176) and Academic Reasoning (IQ 121, #96 of 203). Its lowest-ranked measured dimension is Reliability (IQ 98, #154 of 177).
What is the EQ of hy3?: hy3 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 hy3. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has hy3 been tested on?: hy3 has published results on 9 of the 39 benchmarks behind AI IQ's Composite IQ, including BrowseComp, GPQA Diamond, CritPt, Humanity's Last Exam and SciCode.
How much does hy3 cost?: hy3 costs $0.20 per million input tokens and $0.80 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $0.6543 per million input-plus-output tokens.
Is hy3 open source?: Yes — hy3 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
Who developed hy3?: hy3 was developed by Tencent, based in China.
When was hy3 released?: hy3 was publicly released on April 24, 2026.
How large is the context window of hy3?: hy3 supports a context window of 256,000 tokens — the amount of text it can consider in a single request.
How fast is hy3?: In independent measurements, hy3 generates about 195 tokens per second, with a typical response time of about 4.3 seconds.