RankingEffort benchmarks

Grok 4.7

Compare models

29 measured results · 2 effort settings · 18 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGrok 4.7 · High3 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A75.5Supported
Hard reasoning
A49.1Supported
Coding
A67.2Supported
Human pref
Unknown
Knowledge
A45.9Preliminary
Multimodal
Unknown
Long context
A38.0Preliminary

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Capability score evidence

Grok 4.7 · High: 56.3 overall comparison score. Meets the overall evidence requirements.

Which sources carry the weight?

Artificial Analysis · 100.0% direct comparison weight · independent of the model provider

These shares describe the comparisons entering the fit, not fractions of the final score. Source balancing does not prove independence: overlapping sources may remain correlated.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-09-21.

Both ranks use the same surviving configurations for each removal. The cohort starts with 400 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	Comparison support lost; no rank		165 remaining / 235 lost	Unavailable
ARC Prize	34 → 31	Up 3	400 remaining / 0 lost	61.5 (+5.2 points)
Datacurve	34 → 31	Up 3	400 remaining / 0 lost	53.6 (-2.6 points)
LiveCodeBench	34 → 35	Down 1	400 remaining / 0 lost	56.1 (-0.1 points)
Scale	34 → 34	No movement	400 remaining / 0 lost	57.3 (+1.0 points)
Terminal-Bench	34 → 35	Down 1	400 remaining / 0 lost	57.1 (+0.8 points)
Cognition	34 → 30	Up 4	400 remaining / 0 lost	58.7 (+2.4 points)
LiveBench	34 → 36	Down 2	399 remaining / 1 lost	56.8 (+0.5 points)
SWE-rebench	34 → 33	Up 1	400 remaining / 0 lost	56.5 (+0.2 points)
Epoch AI	34 → 38	Down 4	383 remaining / 17 lost	55.9 (-0.4 points)
Vals AI	34 → 34	No movement	399 remaining / 1 lost	56.2 (-0.1 points)
Provider reports · Anthropic	34 → 34	No movement	400 remaining / 0 lost	56.3 (+0.0 points)
Provider reports · Moonshot AI	34 → 35	Down 1	400 remaining / 0 lost	55.8 (-0.4 points)
Provider reports · Meta	34 → 34	No movement	400 remaining / 0 lost	56.1 (-0.1 points)
Provider reports · Mistral	34 → 34	No movement	400 remaining / 0 lost	56.3 (+0.0 points)
All provider reports	34 → 36	Down 2	400 remaining / 0 lost	55.7 (-0.6 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
aa-lcr	61.5	+5.3
arc-agi	61.2	+5.0
omniscience	59.9	+3.6
briefcase	53.1	-3.1
deepswe	53.7	-2.6
automationbench	53.7	-2.6
hle	58.7	+2.5
frontiercode	58.7	+2.4
critpt	58.6	+2.4
gdpval	54.5	-1.7
ifbench	54.9	-1.3
epoch-game-puzzles	55.0	-1.3
mmmu-pro	57.2	+0.9
scicode	55.5	-0.8
simpleqa	57.0	+0.8
tau-bench	55.6	-0.7
gdp-pdf	55.6	-0.6
livecodebench	55.7	-0.6
gpqa	55.7	-0.6
terminal-bench	56.8	+0.6
arc-agi-3	55.8	-0.5
aime	55.9	-0.3
frontiermath	56.6	+0.3
multi-swe-bench	55.9	-0.3
apex-agents	56.5	+0.3
livebench-language	56.5	+0.2
math500	56.0	-0.2
swe-rebench	56.5	+0.2
swe-atlas-qna	56.0	-0.2
livebench-instructions	56.5	+0.2
enigma-eval	56.0	-0.2
enterprise-ops	56.5	+0.2
gmmlu	56.1	-0.2
swe-atlas-test-writing	56.1	-0.2
vibe-code	56.1	-0.2
harvey	56.1	-0.2
analyst-agent	56.4	+0.2
livebench-reasoning	56.1	-0.1
livebench-coding	56.1	-0.1
vals-legal-research	56.2	-0.1
mmlu-pro	56.2	-0.1
swe-atlas-refactoring	56.2	-0.1
vals-finance-agent	56.2	-0.1
vals-excel-modeling	56.3	+0.0
livebench-math	56.3	+0.0
livebench-data	56.3	+0.0
swe-bench-pro	56.3	+0.0
itbench	56.2	-0.0
vals-code-migration	56.3	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 37 observations, 27 contributing, 0 matched without graph weight and 10 excluded. Counts do not establish rank eligibility.

2 · Benchmark family or source protocol has not been reviewed
5 · The launch table does not establish matching agent harnesses and evaluation settings. Retained as a provider claim; not a new comparable evaluation.
1 · Secondary Artificial Analysis leaderboard report; not an additional independent evaluation. Use the original evaluator for reviewed exact-configuration comparisons.
1 · Secondary Harvey leaderboard report; not an additional independent evaluation. Use the original Vals AI evaluation with its reviewed effort settings.
1 · Secondary Artificial Analysis leaderboard report; retain the original evaluator once, without treating the launch reprint as additional evidence.

Inspect every observation and exclusion →

Benchmark results

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 7,455 results across 241 models · Collected 2026-09-21

Benchmark / source	High	XHigh
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-09-21.	1,643.88 Elo Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	1,657.2 Elo Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	77% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	76.67% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
AutomationBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of task objectives completed with no guardrail violations · Higher is better · Benchmark developed by Zapier · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	63.52% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	65.56% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	18% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	17.71% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
GDP.pdf: All-passArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of attempts where every atomic criterion passed · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	23.2% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	20% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
GDPval-AA v2.1 LeaderboardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Elo rating for performance on real-world work tasks · Anchored to DeepSeek V4.1 Flash (max) at 1,600 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-09-21.	1,693.72 Elo Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	1,695.21 Elo Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	42.31% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	43.14% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-09-21.	30.9 Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	32.03 Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	47.82% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	47.45% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
SciCodeArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	57.75% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	57.41% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
Terminal-Bench v4.0Artificial Analysis · mini-SWE-agent v2.4.6; 66 tasks; pass@1 averaged over 3 repeats; 500 steps; task timeout up to 8 hours Benchmark details Benchmark developed by the Laude Institute, Stanford, and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-21.	24.75% Reported settings Grok 4.7 (high); Model source: https://artificialanalysis.ai/models/grok-4-7-high	25.76% Reported settings Grok 4.7 (xhigh); Model source: https://artificialanalysis.ai/models/grok-4-7
Vals AI · Finance Agent v2Vals AI · Vals FAB v2 shared six-tool harness; two-hour task limit Benchmark details Overall weighted rubric accuracy, averaged over three runs on 450 tasks. Dealbreaker failures and timeouts score zero. Related categories and all-pass scores are not additional votes. Three-model judge panel; private test set. Unit: percent. Source collected 2026-09-21.	Not reported	49.23% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}
Vals AI · Code Migration v1Vals AI · Vals Code Migration shared agent harness Benchmark details Hidden behavior-test pass rate, equally weighted across 40 repositories (30 CLI and 10 COBOL-to-Java). Anti-cheat failures score zero. Code-quality ratings and language splits are not additional votes. Unit: percent. Source collected 2026-09-21.	Not reported	34.71% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}
Vals AI · Legal Research Bench v1Vals AI · Vals Legal Research shared five-tool harness; three-hour task limit Benchmark details All-pass accuracy on 208 private legal research tasks. Every expert rubric check must pass; GPT-5.4 judges responses. Partial-credit scores and practice-area splits are not additional votes. Unit: percent. Source collected 2026-09-21.	Not reported	39.9% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}
Vals AI · Excel Modeling Benchmark v1Vals AI · Vals EMB shared workbook agent harness; 3.5-hour task limit Benchmark details Mean of seven equally weighted task categories. Template submissions use recalculated Excel cell accuracy; scratch submissions use rubric checks judged by GPT-5.4 Mini XHigh. Partial credit does not mean task completion. Unit: percent. Source collected 2026-09-21.	Not reported	54.94% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}
Vals AI · Terminal-Bench 2.1Vals AI · Vals Terminus 2 harness; pass@1 on 89 tasks Benchmark details Overall task pass rate under the shared Terminus 2 configuration. All task tests must pass. Difficulty splits share one family. Fable 5 and Opus 5 mixed-checkpoint fallback runs are excluded. Unit: percent. Source collected 2026-09-21.	Not reported	76.03% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}
Vals AI · Vibe Code Bench v1.1Vals AI · Vals modified OpenHands harness; five hours or 1,000 turns per app Benchmark details Mean app test pass rate on the canonical v1.1 test set. Browser Use executes workflow tests; each test requires at least 90% of its substeps. Automated judging can introduce error. Alternate agents are excluded. Unit: percent. Source collected 2026-09-21.	Not reported	75.86% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":"OpenHands"}
Vals AI · Harvey Legal Agent Benchmark v1Vals AI · Vals Harvey held-out legal file-work protocol; internet disabled Benchmark details Mean task resolution across two judges (GPT-5.5 Medium and Sonnet 4.6), requiring every criterion to pass. Criteria pass rates and practice-area splits do not add votes. Shared Harvey family; Fable 5 fallback run excluded. Unit: percent. Source collected 2026-09-21.	Not reported	19.58% Reported settings {"checkpoint":"grok-4.7","evaluatorModel":"grok/grok-4.7","sourceUpdatedOn":"2026-09-21","reasoning_effort":"xhigh","compute_effort":null,"reasoning":null,"temperature":0.7,"top_p":0.95,"max_output_tokens":null,"verbosity":null,"provider":"SpaceXAI","harness":null}

Grok 4.7

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results