Model Comparator

Compare up to four models using the same criteria.

Llama Nemotron Super 49B v1.5 (Non-reasoning)
NVIDIA
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	Llama Nemotron Super 49B v1.5 (Non-reasoning) NVIDIA	Devstral Small 2 Mistral
Intelligence	7.4	Best: 7.5
Coding	—	29.3
Math	8	Best: 34.3
Agentic	—	2
Speed	35.7 t/s	Best: 142 t/s
TTFT	10.8 s	Best: 2.29 s
Input · $ / 1M tokens	$0.40	—
Output · $ / 1M tokens	$0.40	—
Context window	128K	Best: 256K
Reasoning	No	No
Open weights	Yes	Yes

Intelligence

Llama Nemotron Super 49B v1.5 (Non-reasoning): 7.4
Devstral Small 2: Best: 7.5

Coding

Llama Nemotron Super 49B v1.5 (Non-reasoning): —
Devstral Small 2: 29.3

Math

Llama Nemotron Super 49B v1.5 (Non-reasoning): 8
Devstral Small 2: Best: 34.3

Agentic

Llama Nemotron Super 49B v1.5 (Non-reasoning): —
Devstral Small 2: 2

Speed

Llama Nemotron Super 49B v1.5 (Non-reasoning): 35.7 t/s
Devstral Small 2: Best: 142 t/s

TTFT

Llama Nemotron Super 49B v1.5 (Non-reasoning): 10.8 s
Devstral Small 2: Best: 2.29 s

Input · $ / 1M tokens

Llama Nemotron Super 49B v1.5 (Non-reasoning): $0.40
Devstral Small 2: —

Output · $ / 1M tokens

Llama Nemotron Super 49B v1.5 (Non-reasoning): $0.40
Devstral Small 2: —

Context window

Llama Nemotron Super 49B v1.5 (Non-reasoning): 128K
Devstral Small 2: Best: 256K

Reasoning

Llama Nemotron Super 49B v1.5 (Non-reasoning): No
Devstral Small 2: No

Open weights

Llama Nemotron Super 49B v1.5 (Non-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.

Llama Nemotron Super 49B v1.5 (Non-reasoning)7.4 · 35.7 t/s
Devstral Small 27.5 · 142 t/s