DeepHermes 3 - Llama-3.1 8B Preview

Nous Research

Verified against Artificial Analysis · Jul 13, 2026

Capability

percentile index

Price /M

$0 / $0

in / out

Context

—

Avg score

17.9

6 benchmarks

AA Intelligence

2.3

index

Weights

Open

Released

Feb 13, 2025

About

DeepHermes 3 - Llama-3.1 8B Preview is a open-weights model in the DeepHermes family from Nous Research. Benchmarked on 6 evals, averaging 17.9.

Benchmark scores · 6

avg 17.9 — higher is better

MMLU-Proartificial-analysis · 2026-07

36.5

GPQA Diamondartificial-analysis · 2026-07

MATHartificial-analysis · 2026-07

21.8

SciCodeartificial-analysis · 2026-07

9.1

LiveCodeBenchartificial-analysis · 2026-07

8.5

Humanity's Last Examartificial-analysis · 2026-07

4.3

Details

FamilyDeepHermes

Parameters—

License—

WeightsOpen

ReleasedFeb 13, 2025

History

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$0/M