Skip to content
Leaderboard / Anthropic

Claude Opus 5

Anthropic Max CLOSED Proprietary
Other settings:High · 96
Verdict

Claude Opus 5 is a top-tier model: #12 of 406 by SIscope score. Strongest at chat quality, weaker at agents & tools. On the pricey side: costs more than 94% of priced models.

SIscope score
92
#12 of 406
Human-vote rank
#5
of 419 · score 1508
Price per 1M tokens, blended
$10
$5 in · $25 out
Context window
1M
1,000,000 tokens

Shape

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

CodingAgentsReasoningMathChatInstructions

Category scores

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

Coding96
Agents & tools82
Reasoning & knowledge93
Math94
Chat quality99
Following instructions99

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

Human-vote score over time

Text arena, recent updates.

150015081515Oct 15: 1507.6 (rank #4)Oct 20: 1508.6 (rank #4)Oct 22: 1508.3 (rank #4)Oct 28: 1507.9 (rank #5)Oct 15Oct 28

Coding votes over time

WebDev arena, recent updates.

168016881695Oct 16: 1689.3 (rank #7)Oct 18: 1689.6 (rank #7)Oct 19: 1688.9 (rank #8)Oct 22: 1690.2 (rank #8)Oct 23: 1690.7 (rank #9)Oct 27: 1690.3 (rank #9)Oct 28: 1690.5 (rank #10)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 Diamond94%
PhD-level science multiple choice questions
Better than 96% of 333 tested models. Run Jul 24, 2026.
SimpleQA Verified60%
Short factual questions; rewards not making things up
Better than 84% of 100 tested models. Run Aug 10, 2026.
OTIS Mock AIME 2024-2599%
Competition maths problems in the style of AIME
Better than 92% of 313 tested models. Run Jul 24, 2026.
FrontierMath Tiers 1-386%
Very hard, unpublished research-style maths problems
Better than 90% of 122 tested models. Run Jul 24, 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
#5 of 41915081503–151335,531Oct 28, 2026
WebDev arena
Building web apps, blind human votes
#10 of 15216911684–169719,641Oct 28, 2026
Vision arena
Questions about images, blind human votes
Not on this board yet
Agent arena
Real multi-step agent sessions
#11 of 577.6%6.4–8.8%31,955Nov 2, 2026

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

Price

Input
$5
per 1M tokens
Output
$25
per 1M tokens
Blended
$10
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 (anthropic/claude-opus-5), Nov 4, 2026. Providers can charge differently.