About
Hermes 3 is a generalist language model with many improvements over [Hermes 2](/models/nousresearch/nous-hermes-2-mistral-7b-dpo), including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the...
Related Models
Hermes 4 - Llama-3.1 70B (Reasoning)2025-08-27Hermes 4 - Llama-3.1 405B (Reasoning)2025-08-27Hermes 4 - Llama-3.1 405B (Non-reasoning)2025-08-27Hermes 4 - Llama-3.1 70B (Non-reasoning)2025-08-27Nous: Hermes 4 405B2025-08-26Nous: Hermes 4 70B2025-08-26Nous: Hermes 3 405B Instruct2024-08-16Nous: Hermes 3 405B Instruct (free)2024-08-16
Pricing
Input
$0.70
per 1M tokens
Output
$0.70
per 1M tokens
Blended
$0.70
per 1M tokens
Cheaper than 46% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.70
Monthly
$21.00
vs. Similar Models
Jamba 1.5 LargeQ:0.0
$3.50+400%
Qwen3 8B (Non-reasoning)Q:0.0
$0.18-74%
Qwen3 Omni 30B A3B InstructQ:0.0
$0.43-39%
Qwen3.5 0.8BQ:-0.1
$0.02-97%
Performance
30
tokens/sec
Faster than 2% of models
0.36
seconds
Faster than 96% of models
0.36
seconds
Faster than 98% of models
Market Median
94 tok/s
68% slower
Median TTFT
1.11s
68% faster
Throughput/Dollar
43
tok/s per $/1M
Speed Comparison
Qwen3.5 0.8B
30 tok/s+1%
Qwen3.5 0.8B (Non-reasoning)
30 tok/s+2%
MoonshotAI: Kimi K2 0905
28 tok/s-7%
Context Window
131K
tokens
Larger than 27% of models
Max Output
16K
tokens
13% of context
Benchmarks
MMLU-Pro
57.1%
GPQA Diamond
40.1%
HLE
4.1%
LiveCodeBench
18.8%
SciCode
23.1%
TerminalBench HardNot evaluated
MATH-500
53.8%
AIME
2.3%
AIME 2025Not evaluated
IFBenchNot evaluated
Long Context RecallNot evaluated
Tau2Not evaluated
Market AverageTop Score