Model Comparator

Compare up to four models using the same criteria.

EXAONE 4.5 33B (Reasoning)
LG AI Research
DeepSeek R1 0528 (May '25)
DeepSeek

Differences only

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

Model Comparator
Metrics	EXAONE 4.5 33B (Reasoning) LG AI Research	DeepSeek R1 0528 (May '25) DeepSeek
Intelligence	Best: 13.2	13.1
Coding	23.6	Best: 24.6
Math	—	76
Agentic	—	1.1
Speed	—	31.9 t/s
TTFT	—	1.04 s
Input · $ / 1M tokens	—	$1.35
Output · $ / 1M tokens	—	$3
Context window	Best: 262K	128K
Reasoning	Yes	Yes
Open weights	Yes	Yes

Intelligence

EXAONE 4.5 33B (Reasoning): Best: 13.2
DeepSeek R1 0528 (May '25): 13.1

Coding

EXAONE 4.5 33B (Reasoning): 23.6
DeepSeek R1 0528 (May '25): Best: 24.6

Math

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): 76

Agentic

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): 1.1

Speed

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): 31.9 t/s

TTFT

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): 1.04 s

Input · $ / 1M tokens

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): $1.35

Output · $ / 1M tokens

EXAONE 4.5 33B (Reasoning): —
DeepSeek R1 0528 (May '25): $3

Context window

EXAONE 4.5 33B (Reasoning): Best: 262K
DeepSeek R1 0528 (May '25): 128K

Reasoning

EXAONE 4.5 33B (Reasoning): Yes
DeepSeek R1 0528 (May '25): Yes

Open weights

EXAONE 4.5 33B (Reasoning): Yes
DeepSeek R1 0528 (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.

DeepSeek R1 0528 (May '25)13.1 · $1.76

Not plotted for lack of data on these axes: EXAONE 4.5 33B (Reasoning).