BlogLLM Benchmark Scores as Published, Side by Side
Llama 3.1 70B (pre-trained): Benchmark Scores as Published
Every result the catalog holds for Llama 3.1 70B (pre-trained): 5 tests in 4 areas, from 1 page. All were printed by Meta itself. Each number is as printed; nothing is ranked or rescored.
Developer MetaReleased 2024-07-23Data version 2026103001, as of 2026-10-30
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MATH | 41.6% | Meta, Meta Llama 4 model cardchecked 2026-10-15 | - |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MMLU-Pro | 53.8% | Meta, Meta Llama 4 model cardchecked 2026-10-15 | - |
| MMLU | 79.3% | Meta, Meta Llama 4 model cardchecked 2026-10-15 | - |
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MBPP | 66.4% | Meta, Meta Llama 4 model cardchecked 2026-10-15 | - |
Multilingual
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| TydiQA (F1) | 29.9% | Meta, Meta Llama 4 model cardchecked 2026-10-15 | - |
More from Meta
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.