Shared benchmarks

23
Shared benchmarks for Grok 4.6 and Grok 4.7
BenchmarkGrok 4.6Grok 4.7
  • 65.3%
  • 54.6%
Evaluation details for APEX-Agents Original

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on APEX-Agents Original

Grok 4.7

Record for Grok 4.7 on APEX-Agents Original
  • 84.0%
  • 75.8%
Evaluation details for BrowseComp 130-question subset — Mercor web research agent

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Grok 4.7

  • 20.5%
  • 14.7%
Evaluation details for Chartography October 2026 leaderboard

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Chartography October 2026 leaderboard

Grok 4.7

Record for Grok 4.7 on Chartography October 2026 leaderboard
  • 92.9%
  • 92.0%
Evaluation details for CharXiv Descriptive and reasoning — Mercor single-shot

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

  • 33.4%low
  • 36.1%medium
  • 40.4%high
  • 41.4%xhigh
  • 33.1%low
  • 41.6%medium
  • 43.9%high
  • 46.3%xhigh
Evaluation details for CursorBench 4.0

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on CursorBench 4.0
Record for Grok 4.6 on CursorBench 4.0
Record for Grok 4.6 on CursorBench 4.0
Record for Grok 4.6 on CursorBench 4.0

Grok 4.7

Record for Grok 4.7 on CursorBench 4.0
Record for Grok 4.7 on CursorBench 4.0
Record for Grok 4.7 on CursorBench 4.0
Record for Grok 4.7 on CursorBench 4.0
  • 57.0%
  • 68.0%
Evaluation details for CWE-bench 1 — OpenCode

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on CWE-bench 1 — OpenCode

Grok 4.7

Record for Grok 4.7 on CWE-bench 1 — OpenCode
  • 79.7%
  • 80.3%
Evaluation details for CyberGym Original

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on CyberGym Original

Grok 4.7

Record for Grok 4.7 on CyberGym Original
DeepSWE 1.1Potentially non-equivalent: Evaluator sets differ.
  • 67.0%medium · 1.1 · DataCurve
  • 65.9%high · 1.1 · SpaceXAI
  • 71.0%1.1 · SpaceXAI
Potentially non-equivalent · Evaluation details for DeepSWE 1.1

Evaluator sets differ. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on DeepSWE 1.1
Record for Grok 4.6 on DeepSWE 1.1

Grok 4.7

Record for Grok 4.7 on DeepSWE 1.1
  • 63.7%high
  • 24.8%xhigh
  • 33.3%
Evaluation details for DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps
Record for Grok 4.6 on DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps

Grok 4.7

Record for Grok 4.7 on DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps
  • 53.7%
  • 52.3%
Evaluation details for Finance Agent Benchmark 2 — October 2026 evaluation

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

  • 25.3%
  • 29.5%
Evaluation details for FrontierSWE 2 — October 2026 evaluation

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on FrontierSWE 2 — October 2026 evaluation

Grok 4.7

Record for Grok 4.7 on FrontierSWE 2 — October 2026 evaluation
  • 94.0%high
  • 93.2%xhigh
  • 92.7%
Evaluation details for GPQA Diamond — Epoch AI harness

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness
Record for Grok 4.6 on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness

Grok 4.7

Record for Grok 4.7 on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness
  • 15.8%
  • 12.5%
Evaluation details for HLAB October 2026 evaluation

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Harvey's Legal Agent Benchmark October 2026 evaluation

Grok 4.7

Record for Grok 4.7 on Harvey's Legal Agent Benchmark October 2026 evaluation
  • 49.3%high
  • 48.9%xhigh
  • 56.0%
Evaluation details for SimpleQA Verified Original — Epoch AI harness

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

  • 50.3%
  • 47.0%
Evaluation details for SWE Atlas Codebase QnA — Mercor mini-swe-agent, 250 steps

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

  • 92.7%
  • 84.2%
Evaluation details for SWE-bench Multilingual — Mercor mini-swe-agent, 1000 steps

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

  • 84.6%
  • 79.1%
Evaluation details for SWE-bench Verified — Mercor mini-swe-agent, 1000 steps

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Grok 4.7

  • 78.3%
  • 73.4%
Evaluation details for Terminal-Bench 2.1 — Vals AI Terminus 2

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Terminal-Bench 2.1 — Vals AI Terminus 2

Grok 4.7

Record for Grok 4.7 on Terminal-Bench 2.1 — Vals AI Terminus 2
  • 20.3%
  • 37.6%
Evaluation details for Terminal-Bench 4.0 — Grok Build

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Terminal-Bench 4.0 — Grok Build

Grok 4.7

Record for Grok 4.7 on Terminal-Bench 4.0 — Grok Build
  • 25.3%
  • 6.6%
Evaluation details for Terminal-Bench 4.0 — Mercor mini-swe-agent

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Terminal-Bench 4.0 — Mercor mini-swe-agent

Grok 4.7

Record for Grok 4.7 on Terminal-Bench 4.0 — Mercor mini-swe-agent
  • 17.2%
  • 28.8%
Evaluation details for Terminal-Bench 4.0 — Vals AI mini-swe-agent

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Terminal-Bench 4.0 — Vals AI mini-swe-agent

Grok 4.7

Record for Grok 4.7 on Terminal-Bench 4.0 — Vals AI mini-swe-agent
  • 52.1%
  • 54.9%
Evaluation details for Vals Index 2.1

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

Grok 4.6

Record for Grok 4.6 on Vals Index 2.1

Grok 4.7

Record for Grok 4.7 on Vals Index 2.1
  • 76.2%
  • 86.2%
Evaluation details for Vibe Code Bench 1.1 — OpenHands, October 2026 evaluation

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.