Claude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MPaLM 2 Bison$0.500/MPaLM 2 Unicorn$5.000/MGemma 3 27B$0.270/MGrok 3$6.600/MGrok 2$4.400/MGrok 1.5$8.000/MDeepSeek-V3$0.519/MDeepSeek-V3-0324$0.519/MDeepSeek-R1$1.042/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MPaLM 2 Bison$0.500/MPaLM 2 Unicorn$5.000/MGemma 3 27B$0.270/MGrok 3$6.600/MGrok 2$4.400/MGrok 1.5$8.000/MDeepSeek-V3$0.519/MDeepSeek-V3-0324$0.519/MDeepSeek-R1$1.042/M

nousresearch

EfficientLIVE INDEX

Nous: Hermes 3 70B Instruct

Name: Nous: Hermes 3 70B Instruct
Brand: nousresearch
Price: 0.300000 USD

text->text

The 2024 70B Llama that defined the open-weights chat baseline before Llama 3.3. Still common where hosts haven't upgraded yet.

Profile inherited from upstream Llama 3.1 70B (Hermes 3 base) ↗ — this is a hosted variant of the same open-weights model.

Released Jul 2024Modalities textOfficial model page ↗Pricing source ↗Compare with another model →Estimate monthly cost →

INPUT

$0.300/M

per million input tokens

OUTPUT

$0.300/M

per million output tokens

BLENDED 70/30

$0.300/M

default reference rate

CONTEXT

131K

131,072 tokens

What it's good at

Open weights
128K context
Wide hosted availability

Typical use cases

Self-hosted chat
Fine-tune base
Cost benchmarking

Benchmarks

vs. best public score

Scores inherited from Llama 3.1 70B (Hermes 3 base) — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.

MMLU83%

Multitask academic knowledge across 57 subjects.

GPQA Diamond47%

Graduate-level science questions, "Google-proof".

MATH68%

High-school competition math problems.

HumanEval80%

Python function synthesis from docstrings.

LMArena Elo1248 Elo

Crowd-sourced head-to-head preference Elo rating.

Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources — treat as directional rather than authoritative.