Llama 4 405B.

Metaopen-source405B params

§ 02 · Benchmarks

Every benchmark Llama 4 405B has a recorded score for.

#	Benchmark	Area · Task	Metric	Value	Rank	Date	Source
01	MMLU	Reasoning · Commonsense Reasoning	accuracy	87.8%	#28/64	2025-09-01	source ↗

Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.

§ 03 · Strengths by area

Where Llama 4 405B actually performs.

Reasoning

benchmark

avg rank #28.0

§ 05 · Related models

Other Meta models scored on Codesota.

Llama 3 70B

9 results

Llama 3 (405B, Instruct)

400B total / 17B active (128 experts) params · 7 results

Llama 3.1 70B

4 results

Code Llama 34B

Unknown params · 2 results

ConvNeXt V2 Huge

650M params · 2 results

DeiT-B Distilled

86M params · 2 results

§ 06 · Sources & freshness

Where these numbers come from.

codesota-shadow-mmlu

result

0 of 1 rows marked verified.