Llama 3.1-405B: benchmark scores, price and ranking

Llama 3.1-405B from Meta AI ranks #142 of 743 LLMs on the LLMs Tiger aggregate benchmark score (51.2), based on 3 benchmark results captured up to 2026-10-13.

Organisation: Meta AI
Type: Open-weight / local model
Open weights: Yes
Released: 2024-07-23 (source: epoch)
Aggregate rank: #142 of 743 (score 51.2)

Relative to other models, Llama 3.1-405B is strongest on GSM8K (#2 of 17) and weakest on Epoch ECI (#97 of 204).

Across the 3 benchmarks where at least ten models are scored, Llama 3.1-405B is in the top 10% on 0 and in the top half on 3.

Llama 3.1-405B benchmark results

Llama 3.1-405B benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
GSM8K	96.8%	#2 of 17	HuggingFace GSM8K leaderboard	2026-10-13
MMLU-Pro	61.6%	#57 of 136	HuggingFace MMLU-Pro leaderboard	2026-10-13
Epoch ECI	129	#97 of 204	Epoch AI Benchmarking Hub	2026-10-13

About Llama 3.1-405B

Llama 3.1-405B is an open-weight model that can run on your own hardware from Meta AI. It was released on 2024-07-23, according to Epoch AI. No API price is listed for it yet. We hold 3 benchmark results for it across 2 areas, and none is in the top tenth of its board.

Where Llama 3.1-405B stands

On our aggregate score, Llama 3.1-405B is #142 of 743 ranked models, in the top quarter. Its score is 51.2; the scale is centred on 50, so it sits 1.2 points above the typical model. That beats 81% of the ranked field. Among the 28 ranked Meta AI models it is #5. The next one up is Muse Spark 1.2 (xhigh) at 51.5. The next one down is Muse Wonder (max) at 50.1.

Llama 3.1-405B, benchmark by benchmark

GSM8K puts Llama 3.1-405B at 96.8%, #2 of 17. That beats 88% of the models tested, but 2.8 points short of mimo-v2.5-pro (99.6%). The results either side are mimo-v2.5-pro (99.6%) and granite-4.1-30b (94.2%). It covers grade-school math word problems. The board median is 86.9%, so Llama 3.1-405B is 9.9 points above it. Source: HuggingFace GSM8K leaderboard, captured 2026-10-13.

On MMLU-Pro, Llama 3.1-405B reaches 61.6%. It ranks #57 of 136, ahead of 58% of the field, and 26.5 points behind Darwin-180B-RSI at 88.1%. The results either side are Qwen3-30B-A3B-Base (61.7%) and Nemotron-H-56B-Base-8K (60.5%). It covers 57-subject knowledge + reasoning. The board median is 55.3%, so Llama 3.1-405B is 6.3 points above it. Source: HuggingFace MMLU-Pro leaderboard, captured 2026-10-13.

Llama 3.1-405B manages 129 on Epoch ECI, good for #97 of 204. Qwen 3.8 Max leads at 156, 28 points further up, and 52% of models sit below. The results either side are GPT-4o (Aug 2024) (129) and Mistral Large 2 (Nov 2024) (129). It covers Epoch Capabilities Index: Epoch's own aggregate of ~40 benchmarks. The board median is 127, so Llama 3.1-405B is 1 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Strengths and weaknesses

General knowledge and chat: solid, ahead of 70% of the field on average across 2 benchmarks. Its best board here is GSM8K (#2 of 17) and its weakest is Epoch ECI (#97 of 204).

Release history

The Meta AI model released before it was Meta-Llama-3-8B-Instruct on 2024-04-18, which scores 39.4. The next one was Llama-3.1-405B-Instruct on 2024-07-23, at 43.6.

Reading these numbers

The 3 scores on this page come from 3 sources: Epoch AI Benchmarking Hub (1), HuggingFace GSM8K leaderboard (1), HuggingFace MMLU-Pro leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements. With 3 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

#139 Muse-Glimmer-30B (51.2)
#140 Gemini 2.5 Pro Exp (Mar 2025) (51.2)
#141 GLM-5.1-FP8 (51.2)
#143 Claude 4.5 Sonnet (51.2)
#144 GPT-5 (high) (51.1)
#145 Claude Code + Fable 5 (51.1)