Model Comparator

Compare up to four models using the same criteria.

Hermes 4 - Llama-3.1 405B (Reasoning)
Nous Research
Devstral Small 2
Mistral

Differences only

✓ Best published value among measured models. — Data unavailable. These metrics do not establish an overall ranking.

Model Comparator
Metrics	Hermes 4 - Llama-3.1 405B (Reasoning) Nous Research	Devstral Small 2 Mistral
Intelligence	7.5	7.5
Coding	—	29.3
Math	Best: 69.7	34.3
Agentic	—	2
Speed	39.6 t/s	Best: 142 t/s
TTFT	2.41 s	Best: 2.29 s
Input · $ / 1M tokens	$1	—
Output · $ / 1M tokens	$3	—
Context window	128K	Best: 256K
Reasoning	Yes	No
Open weights	Yes	Yes

Intelligence

Hermes 4 - Llama-3.1 405B (Reasoning): 7.5
Devstral Small 2: 7.5

Coding

Hermes 4 - Llama-3.1 405B (Reasoning): —
Devstral Small 2: 29.3

Math

Hermes 4 - Llama-3.1 405B (Reasoning): Best: 69.7
Devstral Small 2: 34.3

Agentic

Hermes 4 - Llama-3.1 405B (Reasoning): —
Devstral Small 2: 2

Speed

Hermes 4 - Llama-3.1 405B (Reasoning): 39.6 t/s
Devstral Small 2: Best: 142 t/s

TTFT

Hermes 4 - Llama-3.1 405B (Reasoning): 2.41 s
Devstral Small 2: Best: 2.29 s

Input · $ / 1M tokens

Hermes 4 - Llama-3.1 405B (Reasoning): $1
Devstral Small 2: —

Output · $ / 1M tokens

Hermes 4 - Llama-3.1 405B (Reasoning): $3
Devstral Small 2: —

Context window

Hermes 4 - Llama-3.1 405B (Reasoning): 128K
Devstral Small 2: Best: 256K

Reasoning

Hermes 4 - Llama-3.1 405B (Reasoning): Yes
Devstral Small 2: No

Open weights

Hermes 4 - Llama-3.1 405B (Reasoning): Yes
Devstral Small 2: Yes

Each reasoning level remains a separate configuration.

Score and cost

Each line joins the reasoning levels of a compared model; the large point marks the configuration in the table. The top-right corner holds the best score for the lowest cost. Select another point to compare that level.

Hermes 4 - Llama-3.1 405B (Reasoning)7.5 · 39.6 t/s
Devstral Small 27.5 · 142 t/s