Meta AI·released 2025-04-051 source
Llama 4 Maverick benchmark scores: 15 benchmarks tracked. 27% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 15 benchmarks·best result 73.02 on MATH level 5 (reproduced)·4 of 15 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| MATH level 5 | 73.02accuracy (%) | reproduced· optimizedT1 | 2025-04-05Epoch AI |
| Lech Mazur Writing | 62accuracy (%) | unverifiedT1 | 2025-04-05lechmazur/writing Github repository |
| GPQA diamond | 55.98accuracy (%) | reproduced· optimizedT1 | 2025-04-05Epoch AI |
| GeoBench | 52accuracy (%) | unverifiedT1 | 2025-04-05GeoBench leaderboard |
| Fiction.LiveBench | 46.2accuracy (%) | unverifiedT1 | 2025-04-06Fiction.live leaderboard |
| DTBench | 36.45accuracy (%) | unverifiedT1 | 2025-04-05https://conceptualreasoning.ai/dtbench |
| WeirdML | 24.47accuracy (%) | unverifiedT1 | 2025-04-05https://htihle.github.io/weirdml.html |
| OTIS Mock AIME 2024-2025 | 20.48accuracy (%) | reproduced· optimizedT1 | 2025-04-05Epoch AI |
| LMCA | 18.67accuracy (%) | unverifiedT1 | 2025-04-05https://conceptualreasoning.ai/lmca |
| Aider polyglot | 15.6accuracy (%) | unverified· optimizedT1 | 2025-04-05https://aider.chat/docs/leaderboards/ |
| SimpleBench | 13.24accuracy (%) | unverifiedT1 | 2025-04-05SimpleBench Leaderboard |
| ARC-AGI | 4.38accuracy (%) | unverified· optimizedT1 | 2025-04-05https://arcprize.org/leaderboard |
| FrontierMath-2025-02-28-Private | 1.21accuracy (%) | reproduced· optimizedT1 | 2025-04-05Epoch AI |
| HLE | 0.92accuracy (%) | unverifiedT2 | 2025-04-05 |
| ARC-AGI-2 | 0accuracy (%) | unverifiedT2 | 2025-04-05 |
Claims drawn from cited facts, not live model generation.
This model was developed by Meta AI, placing its origin at a major research organization. Its release year of 2025 marks it as a recent vintage.
2 cited facts
This model is scored on 15 tracked benchmarks. It holds no current top score on any of them. An outside evaluation backs these numbers up, so the record can be trusted as independently reproduced rather than taken on the vendor's word.
3 cited facts
On average this model trails the leading score by 58.38 points under the disclosed harness — a wide gap that reads as well behind the leaders rather than at the frontier or merely competitive. It also holds the top score on 0 benchmarks, so there is no category leadership to point to yet.
2 cited facts
5 facts cross-checked across data sources: 5 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks, openrouter_models
Llama 4 Maverick is an AI model developed by Meta AI, released 2025-04-05. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Llama 4 Maverick has recorded scores on 15 benchmarks, each shown with its evidence status.
Llama 4 Maverick has recorded scores on 15 benchmarks — Lech Mazur Writing, GPQA diamond, MATH level 5, OTIS Mock AIME 2024-2025, DTBench, WeirdML, and 9 more. The full table above shows each score with its evidence status.
4 of 15 recorded scores (27%) are independently reproduced rather than self-reported by the lab.
Llama 4 Maverick has 15 tracked claims: 4 independently reproduced, 11 unverified.