Skip to content
Leaderboard / Google

Gemini 2.5 Pro

Google CLOSED Proprietary
Verdict

Gemini 2.5 Pro is a solid mid-pack model: #170 of 406 by SIscope score. Strongest at chat quality, weaker at coding. Mid-range on price.

SIscope score
52
#170 of 406
Human-vote rank
#43
of 419 · score 1458
Price per 1M tokens, blended
$3.44
$1.25 in · $10 out
Context window
1M
1,048,576 tokens

Shape

Each spoke is a category, 0 at the centre, 100 at the edge.

CodingAgentsno data yetReasoningMathChatInstructions

Category scores

Percentiles: 90 means better than about 90% of tested models.

Coding29
Agents & toolsNo data yet–
Reasoning & knowledge77
Math59
Chat quality90
Following instructions85

What goes into each category · Compare with DeepSeek V4.1 Flash

Human-vote score over time

Text arena, recent updates.

145014581465Oct 20: 1457.7 (rank #42)Oct 22: 1457.8 (rank #42)Oct 28: 1457.7 (rank #43)Oct 20Oct 28

Coding votes over time

WebDev arena, recent updates.

122012281235Oct 16: 1226.3 (rank #137)Oct 18: 1226.4 (rank #138)Oct 19: 1226.3 (rank #141)Oct 22: 1226.5 (rank #142)Oct 23: 1226.5 (rank #143)Oct 27: 1226.4 (rank #144)Oct 28: 1226.1 (rank #145)Oct 16Oct 28

Independent test runs

Run by Epoch AI on the same harness for every model. Bars show the share of questions answered correctly.

GPQA Diamond85%
PhD-level science multiple choice questions
Better than 69% of 333 tested models. Run Nov 16, 2025.
OTIS Mock AIME 2024-2585%
Competition maths problems in the style of AIME
Better than 66% of 313 tested models. Run Nov 16, 2025.
FrontierMath Tiers 1-325%
Very hard, unpublished research-style maths problems
Better than 30% of 122 tested models. Run Jun 11, 2026.
SWE-bench Verified58%
Fixing real GitHub issues in Python repositories
Better than 6% of 35 tested models. Run Feb 13, 2026.

Human-vote arenas

People compare two anonymous answers and pick the better one. Scores work like chess ratings.

ArenaRankScoreRange (95%)VotesUpdated
Text arena
Everyday chat, blind human votes
#43 of 41914581455–1460125,307Oct 28, 2026
WebDev arena
Building web apps, blind human votes
#145 of 15212261210–12422,724Oct 28, 2026
Vision arena
Questions about images, blind human votes
#51 of 16712631259–126888,755Oct 19, 2026
Agent arena
Real multi-step agent sessions
Not on this board yet

Source: Arena (formerly LMArena) leaderboard dataset, CC BY 4.0. Fetched Nov 4, 2026.

Price

Input
$1.25
per 1M tokens
Output
$10
per 1M tokens
Blended
$3.44
3 in : 1 out, per 1M
Context window
1M
max tokens per request
Speed
Coming soon
our own tests

List price from the OpenRouter models API (google/gemini-2.5-pro), Nov 4, 2026. Providers can charge differently.