Model Comparator

Compare up to four models using the same criteria.

DeepSeek V4 Pro 0424 (Max)
DeepSeek
K2 Horizon 375B A23B
Institute of Foundation Models

Differences only

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

Model Comparator
Metrics	DeepSeek V4 Pro 0424 (Max) DeepSeek	K2 Horizon 375B A23B Institute of Foundation Models
Intelligence	30.4	Best: 30.5
Coding	59.4	Best: 61.5
Agentic	26.3	Best: 40
Speed	83.8 t/s	Best: 480 t/s
TTFT	Best: 1.79 s	2.72 s
Input · $ / 1M tokens	$0.435	—
Output · $ / 1M tokens	$0.87	—
Context window	Best: 1M	524K
Reasoning	Yes	Yes
Open weights	Yes	Yes
Cost / task	$0.122	Best: $0

Intelligence

DeepSeek V4 Pro 0424 (Max): 30.4
K2 Horizon 375B A23B: Best: 30.5

Coding

DeepSeek V4 Pro 0424 (Max): 59.4
K2 Horizon 375B A23B: Best: 61.5

Agentic

DeepSeek V4 Pro 0424 (Max): 26.3
K2 Horizon 375B A23B: Best: 40

Speed

DeepSeek V4 Pro 0424 (Max): 83.8 t/s
K2 Horizon 375B A23B: Best: 480 t/s

TTFT

DeepSeek V4 Pro 0424 (Max): Best: 1.79 s
K2 Horizon 375B A23B: 2.72 s

Input · $ / 1M tokens

DeepSeek V4 Pro 0424 (Max): $0.435
K2 Horizon 375B A23B: —

Output · $ / 1M tokens

DeepSeek V4 Pro 0424 (Max): $0.87
K2 Horizon 375B A23B: —

Context window

DeepSeek V4 Pro 0424 (Max): Best: 1M
K2 Horizon 375B A23B: 524K

Reasoning

DeepSeek V4 Pro 0424 (Max): Yes
K2 Horizon 375B A23B: Yes

Open weights

DeepSeek V4 Pro 0424 (Max): Yes
K2 Horizon 375B A23B: Yes

Cost / task

DeepSeek V4 Pro 0424 (Max): $0.122
K2 Horizon 375B A23B: Best: $0

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 V4 Pro 0424 (Max)30.4 · $0.122
K2 Horizon 375B A23B30.5 · $0