xAIUnited States· released Jul 2025

Grok 4

Name: Grok 4
Author: xAI

Provisional: 16 comparable tests so far. Ranked once it reaches 6 across 3 counting categories.

xAI, United States. One of 8 models from this lab on the Index.

Third-party publishedLow confidence (thin coverage)

Snapshot September 8, 2026

OpenCharts Index

56.3

Provisional: not enough comparable tests or counting categories to rank yet · 16 comparable tests · 2 counting categories

Drop any single test and the Index lands between 51.9 and 63.7.

4 of 8 hardest tests · share of the frontier

Distance to today's frontier on the hardest tests. Not a measure of intelligence.

What it's known for

Grok 4, in 16 results.

Strongest in reasoning (61.5) and agentic (51.0). It holds the best published result on Fiction.LiveBench. It trails the frontier most in agentic, 49.0% behind on average. No comparable result yet in knowledge. Between 40% and 50% behind the best published results, on average.

Best published result onFiction.LiveBench96.9%

Strongest category

Reasoning

Hard, novel problems: graduate science, abstract puzzles, expert exams.

61.5

GPQA Diamond87.0% · 88% of best
SimpleBench60.5% · 74% of best
ARC-AGI-166.7% · 68% of best
ARC-AGI-216.0% · 17% of best

Mean of 4 of the category's 6 comparable tests. Counts toward the Index.

Best single result

Fiction.LiveBench

Answer questions that need the whole of a very long story, at a 120K-token context length.

96.9%

Best published result of 7 models.

Long documents are the normal case at work. This shows whether a model still holds the thread at the end.

Where it trails

ARC-AGI-2

Infer the hidden rule from a few input and output grid pairs, then apply it to a new grid.

16.0%

83.2% behind the best published result (95.0%) · #38 of 51.

Novel visual puzzles resist memorization, so they measure fluid intelligence rather than recall.

Not measured yet

No comparable result yet in this category. A blank is a blank, never a zero, and it does not lower the Index.

Knowledge

Covered, but not averaged into the Index:

Coding2 of 7 tests · needs 4 · mean 69.9
Math1 of 5 tests · needs 3 · mean 84.0
Multimodalsingle-test category, shown beside the Index · 64.4
Human preferencesingle-test category, shown beside the Index · 73.0
Long contextsingle-test category, shown beside the Index · 96.9

Every score, with its source

See it at work

What Grok 4 was asked to do, and how it did.

Every test it has taken, by category, with the kind of task it faced, who produced the number and where the result landed against the best published one. The examples are original and illustrative, never items from the datasets themselves.

Long context · mean 96.9 across 1 comparable test · a single-test category, shown beside the Index and never averaged into it

Fiction.LiveBenchLeaderboard

Answer questions that need the whole of a very long story, at a 120K-token context length.

96.9%

96.9% of the best (100.0%)
#1 of 7

Source: Fiction.live

Every result is read against the best published one on its test: percent scores as a share of the best above chance, Arena ratings as the expected win rate against the board leader, open-ended values on a log scale. Sitting a harder exam never lowers a model. Every number links to the publisher that produced it.

Head to head

Grok 4 against whoever you pick.

Choose a rival. Every counted test both have taken appears side by side as a share of the best published result on that test, with the category means above. A win is a higher share; a gap under 2 points is a tie, because one item on a 45-problem exam is 2.2 points and publishers' own standard errors are of that order.

Grok 4

provisional·Index 56.3C-

0 : 6

wins · 0 ties · 6 shared

GPT-6 Astra

#1·Index 99.0A+

GPT-6 Astra comes out ahead on 6 of the 6 tests both have taken.

Category means

Reasoning
61.5
99.6
Knowledge
—
100.0
Coding
69.9
97.7
Math
84.0
99.8
Agentic
51.0
98.5
Multimodal
64.4
—
Human preference
73.0
—
Long context
96.9
—

Shared tests · biggest gaps first

ARC-AGI-2Reasoning
16.0%
95.0%
−83.2
APEX-AgentsAgentic
15.2%
46.7%
−66.4
WeirdMLCoding
45.7%
92.9%
−50.8
ARC-AGI-1Reasoning
66.7%
98.5%
−32.3
Mock AIME 2024–2025Math
84.0%
100.0%
−16.0
GPQA DiamondReasoning
87.0%
95.8%
−12.4

Bars are distances to the best published result on each test (percent scores above chance, Arena ratings as win rate against the leader, open-ended values on a log scale). Raw values as the publishers report them.

Open GPT-6 Astra

Scores

Every test, every source

The best published run per benchmark, the raw value as the publisher reports it, who produced the number, how close it comes to the best published result on that test, and where the model sits among every model scored on it. Percent scores are read above chance, Arena ratings as a win rate against the board leader, open-ended values on a log scale. A test with too few models to compare against is shown but not counted.

Category means

Reasoning61.5 · 4/6
Knowledge—
Coding69.9 · 2/7
Math84.0 · 1/5
Agentic51.0 · 6/8
Multimodalbeside64.4 · 1/1
Human preferencebeside73.0 · 1/1
Long contextbeside96.9 · 1/1

Mean · tests taken of the category's comparable tests. A faded category either rests on fewer than half of them or is a single-test category (marked “beside”), which is shown next to the Index and never averaged into it.

GPQA DiamondEpoch-runReasoning

87.0%

87.6% of the best (95.8%), above chance
42th percentile of 78 · n = 198

Source: Epoch AI — AI Benchmarking Hub

ARC-AGI-2LeaderboardReasoning

16.0%

16.8% of the best (95.0%)
26th percentile of 51 · n = 120

Source: Epoch AI — AI Benchmarking Hub

ARC-AGI-1LeaderboardReasoning

66.7%

67.7% of the best (98.5%)
28th percentile of 51 · n = 100

Source: Epoch AI — AI Benchmarking Hub