GPQA Diamond ↗89.33%
Grok 4.20 · GPQA diamond · Best score (across scorers) · % ↗
xAI / MODEL EVIDENCE
Grok 4.20
20 benchmarks · 21 observations · 20 sources · released 2026-02-17
Scores at a glance
20 benchmarks · large dot: Grok 4.20; small dots: closest models where reported
Grok 4.20 · FrontierMath-Tiers-1-3-v2-Private · Best score (across scorers) · % ↗
Grok 4.20 · FrontierMath-Tier-4-v2-Private · Best score (across scorers) · % ↗
Grok 4.20 · OTIS Mock AIME 2024-2025 · Best score (across scorers) · % ↗
SimpleQA Verified ↗30.2%
Grok 4.20 · SimpleQA Verified · Best score (across scorers) · % ↗
Grok 4.20 · Chess Puzzles · Best score (across scorers) · % ↗
ARC-AGI-1 ↗89.5%
Grok 4.20 · ARC-AGI · Score · % ↗
Terminal-Bench 2.0 ↗57.3%
Grok 4.20 · Terminal Bench · Accuracy mean · % ↗
WeirdML (v2) ↗52.26%
Grok 4.20 · WeirdML · Accuracy · % ↗
ARC-AGI-2 ↗65.14%
Grok 4.20 · ARC-AGI-2 · Score · % ↗
CL-bench ↗22.2%
Grok 4.20 · CL-bench · Overall · % ↗
CL-bench Life ↗11.9%
Grok 4.20 · CL-bench Life · Overall · % ↗
ProofBench ↗14%
Grok 4.20 · ProofBench · Accuracy · % ↗
LMCA ↗38.66%
Grok 4.20 · LMCA · Score · % ↗
DTBench ↗90.13%
Grok 4.20 · DTBench · Accuracy · % ↗
Text Arena (Coding) ↗1373.76 Elo
Grok 4.20 · WebDev Arena · Elo ↗
Grok 4.20 · Blueprint-Bench 2 · Score · % ↗
Epoch Capabilities Index (ECI) ↗152.11 ECI
Grok 4.20 · General ECI · 26 October 2026 fit ↗
Grok 4.20 · HiL-Bench (Human-in-Loop Benchmark) · Score ↗
Vending-Bench 2 ↗4662.84667 USD
Grok 4.20 · End-of-year balance ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
5 of 7 shared benchmark comparisons
GLM-5.3-Flash →Z.ai
5 of 7 shared benchmark comparisons
6 of 7 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- Blueprint-Bench 2 | Andon Labs ↗
- ARC Prize - Leaderboard ↗
- Arena leaderboard — WebDev evaluation ↗
- DTBench — Conceptual Reasoning Index ↗
- LMCA — Conceptual Reasoning Index ↗
- ARC-AGI-2 ↗
- Chess Puzzles ↗
- CL-bench ↗
- CL-bench Life ↗
- FrontierMath Tier 4 (v2) ↗
- FrontierMath Tiers 1-3 (v2) ↗
- GPQA Diamond ↗
- OTIS Mock AIME 2024-2025 ↗
- SimpleQA Verified ↗
- Vending Bench 2 — methodology and published results ↗
- AI benchmarks & capabilities ↗
- WeirdML ↗
- HiL-Bench (Human-in-Loop Benchmark) — evaluation methodology and results ↗
- Terminal-Bench 2.0 leaderboard ↗
- ProofBench — evaluation methodology and results ↗
Names used in source tables
Grok 4.20