Model Comparator

Compare up to four models using the same criteria.

Mistral Medium 3
Mistral
Granite 4.2 3B
IBM

Differences only

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

Model Comparator
Metrics	Mistral Medium 3 Mistral	Granite 4.2 3B IBM
Intelligence	9	Best: 9.1
Coding	—	17.5
Math	30.3	—
Agentic	—	0.9
Speed	45.1 t/s	Best: 219 t/s
TTFT	2.24 s	Best: 0.53 s
Input · $ / 1M tokens	$0.40×13.3 relative to the lowest price among the compared models	Best: $0.030
Output · $ / 1M tokens	$2×16.7 relative to the lowest price among the compared models	Best: $0.12
Context window	128K	Best: 131K
Reasoning	No	Yes
Open weights	No	Yes
Cost / task	—	$0.0060

Intelligence

Mistral Medium 3: 9
Granite 4.2 3B: Best: 9.1

Coding

Mistral Medium 3: —
Granite 4.2 3B: 17.5

Math

Mistral Medium 3: 30.3
Granite 4.2 3B: —

Agentic

Mistral Medium 3: —
Granite 4.2 3B: 0.9

Speed

Mistral Medium 3: 45.1 t/s
Granite 4.2 3B: Best: 219 t/s

TTFT

Mistral Medium 3: 2.24 s
Granite 4.2 3B: Best: 0.53 s

Input · $ / 1M tokens

Mistral Medium 3: $0.40×13.3 relative to the lowest price among the compared models
Granite 4.2 3B: Best: $0.030

Output · $ / 1M tokens

Mistral Medium 3: $2×16.7 relative to the lowest price among the compared models
Granite 4.2 3B: Best: $0.12

Context window

Mistral Medium 3: 128K
Granite 4.2 3B: Best: 131K

Reasoning

Mistral Medium 3: No
Granite 4.2 3B: Yes

Open weights

Mistral Medium 3: No
Granite 4.2 3B: Yes

Cost / task

Mistral Medium 3: —
Granite 4.2 3B: $0.0060

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.

Mistral Medium 39 · $0.80
Granite 4.2 3B9.1 · $0.052