Llama 4 Scout · GPQA diamond · Best score (across scorers) · % ↗
1 more reported result
46.97 % · Llama 4 Scout · GPQA Diamond · Accuracy · Vals AI · GPQA Diamond ↗
Meta / MODEL EVIDENCE
28 benchmarks · 39 observations · 30 sources · released 2025-04-05
28 benchmarks · large dot: Llama 4 Scout; small dots: closest models where reported
Llama 4 Scout · GPQA diamond · Best score (across scorers) · % ↗
46.97 % · Llama 4 Scout · GPQA Diamond · Accuracy · Vals AI · GPQA Diamond ↗
Llama 4 Scout · FrontierMath-2025-02-28-Private · Best score (across scorers) · % ↗
Llama 4 Scout · MATH level 5 · Best score (across scorers) · % ↗
Llama 4 Scout · OTIS Mock AIME 2024-2025 · Best score (across scorers) · % ↗
Llama 4 Scout · ARC-AGI · Score · % ↗
Llama 4 Scout · Fiction.LiveBench · 16k token score · % ↗
Llama 4 Scout · ARC-AGI-2 · Score · % ↗
Llama 4 Scout · CritPt · Accuracy · % ↗
Llama 4 Scout · LMCA · Score · % ↗
Llama 4 Scout · DTBench · Accuracy · % ↗
Llama 4 Scout · SciCode · Score · % ↗
Llama 4 Scout · SpatialViz-Bench · Overall score · % ↗
Llama 4 Scout · General ECI · 26 October 2026 fit ↗
Llama 4 Scout · LegalBench · Accuracy ↗
Llama 4 Scout · CorpFin · Accuracy ↗
Llama 4 Scout · MortgageTax · Accuracy ↗
55.19 % · Llama 4 Scout · TaxEval · Accuracy · Vals AI · TaxEval ↗
Llama 4 Scout · MedCode · Accuracy ↗
Llama 4 Scout · MedScribe · Accuracy ↗
Llama 4 Scout · MedQA · Accuracy ↗
Llama 4 Scout · AIME · Accuracy ↗
Llama 4 Scout · MATH 500 · Accuracy ↗
Llama 4 Scout · MGSM · Accuracy ↗
Llama 4 Scout · MMLU Pro · Accuracy ↗
Llama 4 Scout · MMMU Pro · Accuracy ↗
Llama 4 Scout · SAGE · Accuracy ↗
Llama 4 Scout · LiveCodeBench · Accuracy ↗
Llama 4 Scout · TextQuests · No clues · game progress · source 500-step protocol ↗
7.7% · Llama 4 Scout · TextQuests · With clues · game progress · source 500-step protocol · TextQuests: How Good are LLMs at Text-Based Video Games? ↗
Llama 4 Scout · CyberSecEval · CyberSOCEval · malware analysis · Overall accuracy · CyberSecEval 4 · CyberSOCEval defensive-analysis comparison ↗
26% · Llama 4 Scout · CyberSecEval · CyberSOCEval · malware analysis · Easy accuracy · CyberSecEval 4 · CyberSOCEval defensive-analysis comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
16% · Llama 4 Scout · CyberSecEval · CyberSOCEval · malware analysis · Medium accuracy · CyberSecEval 4 · CyberSOCEval defensive-analysis comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
0% · Llama 4 Scout · CyberSecEval · CyberSOCEval · malware analysis · Hard accuracy · CyberSecEval 4 · CyberSOCEval defensive-analysis comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
43.76% · Llama 4 Scout · CyberSecEval · CyberSOCEval · threat intelligence · image input · Overall accuracy · CyberSecEval 4 · CyberSOCEval image-input comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
62% · Llama 4 Scout · CyberSecEval · CyberSOCEval · threat intelligence · image input · NSA reports · CyberSecEval 4 · CyberSOCEval image-input comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
47% · Llama 4 Scout · CyberSecEval · CyberSOCEval · threat intelligence · image input · CrowdStrike reports · CyberSecEval 4 · CyberSOCEval image-input comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
37% · Llama 4 Scout · CyberSecEval · CyberSOCEval · threat intelligence · image input · IC3 reports · CyberSecEval 4 · CyberSOCEval image-input comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
38% · Llama 4 Scout · CyberSecEval · CyberSOCEval · threat intelligence · image input · CISA reports · CyberSecEval 4 · CyberSOCEval image-input comparison · CyberSOCEval: Benchmarking LLMs Capabilities for Malware Analysis and Threat Intelligence Reasoning ↗
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
12 of 17 shared benchmark comparisons
13 of 17 shared benchmark comparisons
11 of 17 shared benchmark comparisons
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.
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
Wang, Siting, Pei, Minnan, Sun, Luoyang, Deng, Cheng, Li, Yuchen, Shao, Kun, Tian, Zheng, Zhang, Haifeng, Wang, Jun
Deason, Lauren, Bali, Adam, Bejean, Ciprian, Bolocan, Diana, Crnkovich, James, Croitoru, Ioana, Durai, Krishna, Midler, Chase, Miron, Calin, Molnar, David, Moon, Brad, Ostarcevic, Bruno, Peltea, Alberto, Rosenberg, Matt, Sandu, Catalin, Saputkin, Arthur, Shah, Sagar, Stan, Daniel, Szocs, Ernest, Wan, Shengye, Whitman, Spencer, Krasser, Sven, Saxe, Joshua
Llama 4 Scout