Model Horizon
DashboardModelsCompareBenchmarks
© 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
SpecClaude Haiku 4.5
Anthropic
Gemini 3.1 Pro
Google
Tiermidfrontier
Release Date2025-10-012026-02-19
Context Window
200K
1049K
Max Output
64K
66K
Input Price (/1M)
$1
$2
Output Price (/1M)
$5
$12
Arena Elo
1,404
1,500
MMLU
82%
92.6%
GPQA
55%
94.3%
MATH—
96.8%
HumanEval
90%
94.6%
SWE-bench—
80.6%
AIME
28%
91.2%
SimpleQA
19%
79.6%
Capabilitiesvision, tool-use, code, extended-thinkingvision, tool-use, code, reasoning, agentic, audio

Benchmark Comparison

Claude Haiku 4.5
Gemini 3.1 Pro
Arena Elo
1,404
1,500
MMLU
82%
92.6%
GPQA
55%
94.3%
MATH
N/A
96.8%
HumanEval
90%
94.6%
SWE-bench
N/A
80.6%
AIME
28%
91.2%
SimpleQA
19%
79.6%
Cost Estimate

Cost Calculator

Claude Haiku 4.5
$9.00/mo$0.3000/day
Gemini 3.1 Pro
$21.00/mo$0.7000/day
Try These Models
Try Claude Haiku 4.5Try Gemini 3.1 Pro