Model Horizon
DashboardModelsCompare
© 2026 Model Horizon
About|Terms
SYS.v0.1.0
Skip to content
  1. Home
  2. /Compare

Compare Models

Side-by-side comparison of AI model specs, benchmarks, and pricing.

Swap
Specifications
SpecDeepSeek V3.2
DeepSeek
Claude Sonnet 4.6
Anthropic
Tieropenfrontier
Release Date2025-12-012026-02-17
Context Window
131K
200K
Max Output
131K
64K
Input Price (/1M)
$0.28
$3
Output Price (/1M)
$0.42
$15
Arena Elo
1,419
1,470
MMLU
85.7%
91%
GPQA
79.9%
88%
MATH—
97.8%
HumanEval
92%
96%
SWE-bench
52%
70.3%
Capabilitiestool-use, codevision, tool-use, code, extended-thinking

Benchmark Comparison

DeepSeek V3.2
Claude Sonnet 4.6
Arena Elo
1,419
1,470
MMLU
85.7%
91%
GPQA
79.9%
88%
MATH
N/A
97.8%
HumanEval
92%
96%
SWE-bench
52%
70.3%
Cost Estimate

Cost Calculator

DeepSeek V3.2
$1.05/mo$0.0350/day
Claude Sonnet 4.6
$27.00/mo$0.9000/day