Meta AI·released 2024-07-14✓ 2 sources
Llama 3.1-8B benchmark scores: 11 benchmarks tracked. 9% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 11 benchmarks·best result 82.4 on GSM8K (self-reported)·1 of 11 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| GSM8K | 82.4accuracy (%) | self-reported· optimizedT1 | 2024-07-23Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone |
| PIQA | 62.4accuracy (%) | self-reported· optimizedT1 | 2024-07-23Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone |
| MMLU | 41.47accuracy (%) | unverified· optimizedT1 | 2024-07-23Stanford CRFM Leaderboard |
| MATH level 5 | 22.88accuracy (%) | reproduced· optimizedT1 | 2024-07-23Epoch AI |
| DTBench | 18.22accuracy (%) | unverifiedT1 | 2024-07-23https://conceptualreasoning.ai/dtbench |
| Balrog | 15.1accuracy (%) | unverifiedT1 | 2024-07-23https://balrogai.com/ |
| LMCA | 6.32accuracy (%) | unverifiedT1 | 2024-07-23https://conceptualreasoning.ai/lmca |
| GPQA diamond | 2.61accuracy (%) | unverifiedT2 | 2024-07-23 |
| WeirdML | 1.73accuracy (%) | unverifiedT1 | 2024-07-23https://htihle.github.io/weirdml.html |
| OTIS Mock AIME 2024-2025 | 1.57accuracy (%) | unverifiedT2 | 2024-07-23 |
| Chess Puzzles | 0accuracy (%) | unverifiedT2 | 2024-07-23 |
Claims drawn from cited facts, not live model generation.
This model was developed by Meta AI and released in 2024, making it a recent entry from a major developer. At 8.03 billion parameters, it is a compact model whose scale favors efficiency and lower cost over the greater headroom and expense of larger systems.
3 cited facts
The model is scored on 11 tracked benchmarks. It holds no current top score on any of them. Its record can be taken at face value, since outside evaluation has independently reproduced its results.
3 cited facts
On average this model trails the leading score by 64.26 points across its benchmarks, a wide gap under the disclosed harness that reads as well behind the front of the pack rather than competitive. It holds the top score on none of its benchmarks, so there is no genuine category leadership to point to yet.
2 cited facts
2 facts cross-checked across data sources: 1 corroborated, 1 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks, huggingface_models
Llama 3.1-8B is an AI model developed by Meta AI, released 2024-07-14. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Llama 3.1-8B has recorded scores on 11 benchmarks, each shown with its evidence status.
Llama 3.1-8B has recorded scores on 11 benchmarks — PIQA, GPQA diamond, MATH level 5, OTIS Mock AIME 2024-2025, DTBench, WeirdML, and 5 more. The full table above shows each score with its evidence status.
1 of 11 recorded scores (9%) are independently reproduced rather than self-reported by the lab.
Llama 3.1-8B has 11 tracked claims: 1 independently reproduced, 2 self-reported, 8 unverified.