TencentReleased Nov 4, 2026
NewProprietary

Hy4

Strongest in Computer Use (#17 of 186) and Reliability (#28 of 188). Five of seven dimensions are estimated, because no results are published yet.

Composite IQ ?
118
#61 of 186 ranked models · ahead of 68%
China (Other)'s #3 model, behind Ling 3.1 Flash (120)
Effective cost
$5.01/1M
$0.83 in · $2.50 out
Speed
Not measured
No speed data yet
Context
1Mtokens
1,000,000 tokens in
Evidence
2of 7 dims
2 of 39 benchmarks

IQ profile

Seven dimensions, averaged equally into Composite IQ
Compare shape with
100 avg human130160● Abstract92#108 · est.● Math135#24 · est.● Academic100#183 · est.● Coding137#29 · est.● Computer Use133#17● Reliability125#28● Knowledge Work102#117 · est.
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
92estimated
#108 / 221
0/5
-40.5
135estimated
#24 / 186
0/5
-14.5
100estimated
#183 / 223
0/6
-35
137estimated
#29 / 187
0/6
-14
133measured
#17 / 186
1/5
-13.5
125measured
#28 / 188
1/6
-10.5
102estimated
#117 / 186
0/6
-30
Hy495% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

2 of 39 Composite IQ benchmarks have a published result for Hy4. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 92 · estimatedNo published results0/5
No published results yet for Abstract Reasoning. The 92 shown is estimated from Hy4'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 135 · estimatedNo published results0/5
No published results yet for Mathematical Reasoning. The 135 shown is estimated from Hy4'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 100 · estimatedNo published results0/6
No published results yet for Academic Reasoning. The 100 shown is estimated from Hy4's other dimensions and will be replaced as results come in. How estimates work
Humanity's Last ExamGPQA DiamondCritPtSciCodeMMLU-ProMMMU-Pro
Programmatic Reasoning IQ 137 · estimatedNo published results0/6
No published results yet for Programmatic Reasoning. The 137 shown is estimated from Hy4's other dimensions and will be replaced as results come in. How estimates work
LiveCodeBenchIOI V2Terminal-Bench 4.0ProgramBenchSWE-rebenchMirrorCode
Computer Use IQ 133 · measuredBest: #20 of 77 on Arena.ai Agent Arena1/5
BenchmarkScorePercentileRankLeader · source
Arena.ai Agent Arena
Arena.ai Agent Arena · Nov 19
6.1
#20 / 77
Leader fable-5.2 · 27.4
Arena.ai Agent Arena · Nov 19
Not yet reported
BrowseCompOSWorld-VerifiedMCP AtlasAgents' Last Exam
Reliability IQ 125 · measuredBest: #33 of 96 on BullshitBench v21/6
BenchmarkScorePercentileRankLeader · source
BullshitBench v2
Full source leaderboard reconciliation · Nov 17
57.0%
#33 / 96
Leader opus-4.8 · 94.0%
Full source leaderboard reconciliation · Nov 17
Not yet reported
SimpleQA VerifiedaaOmniscienceIFBenchAA Long Context Reasoning v1.1FACTS Grounding
Knowledge Work IQ 102 · estimatedNo published results0/6
No published results yet for Knowledge Work. The 102 shown is estimated from Hy4'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
Tencent
Access
Available through Tencent Cloud TokenHub and Tencent products; also offered through select third-party API platforms.
Weights
ProprietaryHosted access only
Inputs
textimage
Context
1,000,000 tokens
API price
$0.83 in · $2.50 outper 1M tokens · full pricing

What Tencent says

  • Released November 4 as the distinct production successor to Hy4 Preview, with image understanding, reasoning, tool use and computer-use capabilities.
  • Tencent Cloud lists the hy4 API ID and a one-million-token context window.

Limitations

  • Tencent announced plans to release Hy4 weights and a system card, but an exact downloadable production checkpoint was not verified as of November 17.
  • The listed US-dollar rates are Tencent Cloud Singapore standard prices; a temporary 50% TokenHub promotion runs through December 31.
  • Publisher-reported evaluations are not imported as independent benchmark results.

Publisher statements, not AI IQ measurements.

About this score

How much of Hy4's IQ is measured, and how the rest is estimated.

2/ 39benchmark results
2/ 7dimensions measured
5estimated
0/5Abstract
0/5Math
0/6Academic
0/6Coding
1/5Computer Use
1/6Reliability
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 and Knowledge Work have no published results yet, so their IQ is estimated from Hy4's other dimensions. Missing benchmarks can also draw on its predecessor, Hy4; 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.

Read the scoring methodology →

Compare Hy4

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

Hy4Comparison model
Browse all comparisons →

Frequently asked questions

Quick answers about Hy4, 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 Hy4?
Hy4 has an estimated IQ of 118, which ranks #61 out of 186 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 Hy4 best at?
Among the dimensions with published benchmark results, Hy4 ranks highest in Computer Use (IQ 133, #17 of 186) and Reliability (IQ 125, #28 of 188).
What is the EQ of Hy4?
Hy4 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 Hy4. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has Hy4 been tested on?
Hy4 has published results on 2 of the 39 benchmarks behind AI IQ's Composite IQ, including Arena.ai Agent Arena and BullshitBench v2.
How much does Hy4 cost?
Hy4 costs $0.834 per million input tokens and $2.50 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $5.01 per million input-plus-output tokens.
Is Hy4 open source?
No — Hy4 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.
Who developed Hy4?
Hy4 was developed by Tencent, based in China.
When was Hy4 released?
Hy4 was publicly released on November 4, 2026.
How large is the context window of Hy4?
Hy4 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
How fast is Hy4?
AI IQ doesn't have independent speed measurements for Hy4 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.