MMODELYST
Models/Hermes 4 - Llama-3.1 70B
MDL

Hermes 4 - Llama-3.1 70B

Nous Research

Verified against Artificial Analysis · Jul 13, 2026
Capability
55
percentile index
Price /M
$0.13 / $0.4
in / out
Context
Avg score
39.2
10 benchmarks
AA Intelligence
10.0
index
Speed
97 tok/s
TTFT 0.6s
Released
Aug 27, 2025
About

Hermes 4 - Llama-3.1 70B is a open-weights model in the Hermes family from Nous Research. Benchmarked on 10 evals, averaging 39.2.

Benchmark scores · 10
avg 39.2 — higher is better
MMLU-Pro
81.1
GPQA Diamond
69.9
AIME
68.7
LiveCodeBench
65.3
SciCode
34.1
IFBench
31.3
τ²-bench
22.5
Humanity's Last Exam
7.9
AA Long-Context Reasoning
6.7
Terminal-Bench Hard
4.5
Details
PublisherNous Research
FamilyHermes
Parameters
License
WeightsOpen
ReleasedAug 27, 2025
Performance & price
Output speed97 tok/s
Latency (TTFT)0.6s
First answer21s
Price in / out$0.13 / $0.4 per 1M
Blended price$0.198 / 1M
Math index68.7

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time

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

API providers
Artificial Analysis$0.4/M