supermarket

Language

Grok 4.5

By SpaceXAI (xAI), in the Grok family.

Released
8 Jul 2026
Status
Legacy
Context
500K
Max output
—
Input per 1M tokens
$2.00
Output per 1M tokens
$6.00
Open weights
No
License
proprietary
Parameters
—
Knowledge cutoff
—
Reasoning
Yes
Modalities
text, image → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Knowledge & reasoning49.9163 of 4
Coding & software engineering48.1Not ranked2 of 5
Factuality & hallucination48.3261 of 1
Human preference53.3214 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Chess PuzzlesPuzzles solved36.0%Independenthigh8 Jul 2026Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)96.5%Third partyhigh—https://conceptualreasoning.ai/dtbench via Epoch AI
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.452Third partyhigh—https://conceptualreasoning.ai/lmca via Epoch AI
SimpleBenchScore (AVG@5)70.0%Third partyunknown—https://lmcouncil.ai/benchmarks via Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy93.4%Independenthigh8 Jul 2026Epoch AI
Surface Evolver BenchMean score (partial credit)74.4%Third partyhigh—https://yhenon.github.io/surface-evolver-llm-eval/ via Epoch AI

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierMath Tier 4Accuracy24.4%Independenthigh9 Jul 2026Epoch AI
FrontierMath Tiers 1-3Accuracy57.2%Independenthigh9 Jul 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy97.8%Independenthigh8 Jul 2026Epoch AI

Coding & software engineering

Coding & software engineering results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
DeepSWEPass@153.8%Third partyhigh—https://deepswe.datacurve.ai/ via Epoch AI
FrontierCodeMain score (Mean@5)42.4%Third partyunknown—https://cognition.com/frontiercode via Epoch AI
PostTrainBenchAverage score23.4%Source not statedhigh—Epoch AI via Epoch AI
WeirdML (v2)Average accuracy46.4%Third partyunknown—https://htihle.github.io/weirdml.html via Epoch AI

Professional knowledge work

Professional knowledge work results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
APEX-AgentsPass@1 (all rubric criteria met)56.2%Source not statedunknown—Epoch AI via Epoch AI

Vision & multimodal understanding

Vision & multimodal understanding results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Furniture AssemblyScore22.5%Independenthigh24 Sep 2026Epoch AI

Factuality & hallucination

Factuality & hallucination results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
SimpleQA VerifiedAccuracy (% correct)48.3%Independenthigh27 Aug 2026Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1552Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes11302Independent—25 Sep 2026Arena
Document Arena (arena.ai)Arena score (Elo scale)1464Independent—13 Sep 2026Arena
Document Arena (arena.ai)Votes4965Independent—13 Sep 2026Arena
Search Arena (arena.ai)Arena score (Elo scale)1202Independent—24 Aug 2026Arena
Search Arena (arena.ai)Votes31505Independent—24 Aug 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1465Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes37535Independent—25 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1279Independent—28 Sep 2026Arena
Vision Arena (arena.ai)Votes10681Independent—28 Sep 2026Arena

Other Grok versions

Sources