Google DeepMind • Listed June 2025 • Multimodal Foundation

Gemini 2.5 Pro

Arena #86 • score 1445 ±2

Proprietary model from Google DeepMind, listed June 2025. Arena text score 1445 (#86 of 414 entries). $1.25 input / $10.00 output per 1M tokens. 1M tokens context.

Parameters
Not disclosed
Context Window
1M tokens
Input / 1M tokens
$1.25
Output / 1M tokens
$10.00

Benchmark Results

  • Arena Text Elo: 1445 (#86 of 414 on arena.ai · ±2 · 125,379 votes · variant: gemini-2.5-pro)
  • SWE-bench Verified: 53.6% (mini-SWE-agent (bash only) · run 'Gemini 2.5 Pro (2025-05-06)' · 2025-07-26)
  • AA Intelligence Index: 16.1 (Artificial Analysis v4.3 · best setting: Gemini 2.5 Pro · checked 2026-10-11)
  • AA Coding Index: 33.3 (Artificial Analysis v4.3 · best setting: Gemini 2.5 Pro · checked 2026-10-11)
  • AA Agentic Index: 1.6 (Artificial Analysis v4.3 · best setting: Gemini 2.5 Pro · checked 2026-10-11)

Sources