Model Comparator

Compare up to four models using the same criteria.

Nova 2.0 Lite (Non-reasoning)
Amazon
Devstral Small (May '25)
Mistral

Differences only

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

Model Comparator
Metrics	Nova 2.0 Lite (Non-reasoning) Amazon	Devstral Small (May '25) Mistral
Intelligence	8.7	8.7
Math	33.7	—
Speed	167 t/s	—
TTFT	1.18 s	—
Input · $ / 1M tokens	$0.33	—
Output · $ / 1M tokens	$2.75	—
Context window	Best: 1M	256K
Reasoning	No	No
Open weights	No	Yes

Intelligence

Nova 2.0 Lite (Non-reasoning): 8.7
Devstral Small (May '25): 8.7

Math

Nova 2.0 Lite (Non-reasoning): 33.7
Devstral Small (May '25): —

Speed

Nova 2.0 Lite (Non-reasoning): 167 t/s
Devstral Small (May '25): —

TTFT

Nova 2.0 Lite (Non-reasoning): 1.18 s
Devstral Small (May '25): —

Input · $ / 1M tokens

Nova 2.0 Lite (Non-reasoning): $0.33
Devstral Small (May '25): —

Output · $ / 1M tokens

Nova 2.0 Lite (Non-reasoning): $2.75
Devstral Small (May '25): —

Context window

Nova 2.0 Lite (Non-reasoning): Best: 1M
Devstral Small (May '25): 256K

Reasoning

Nova 2.0 Lite (Non-reasoning): No
Devstral Small (May '25): No

Open weights

Nova 2.0 Lite (Non-reasoning): No
Devstral Small (May '25): 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.

Nova 2.0 Lite (Non-reasoning)8.7 · $0.935

Not plotted for lack of data on these axes: Devstral Small (May '25).