Skip to main content

Claude Sonnet 4.5 vs Claude Sonnet 5.5

SpecClaude Sonnet 4.5Claude Sonnet 5.5
Respan Index31.3 (#86)73.4 (#6)
Blended price per 1M$6$4
Context window200K1M
Input / output per 1M$3 / $15$2 / $10
Max output64K128K
Inputtext, visiontext, vision
Released2025-09-292026-09-28

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryClaude Sonnet 4.5Claude Sonnet 5.5
Coding13.1 Claude Sonnet 4.578.8 Claude Sonnet 5.5
Agents41.7 Claude Sonnet 4.580.1 Claude Sonnet 5.5
Reasoning27.5 Claude Sonnet 4.579.0 Claude Sonnet 5.5
Knowledge & language31.2 Claude Sonnet 4.548.6 Claude Sonnet 5.5
Human preference50.9 Claude Sonnet 4.560.2 Claude Sonnet 5.5

How the Respan Index works

Benchmarks both report

Claude Sonnet 5.5 leads on 14 of the 14 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkClaude Sonnet 4.5Claude Sonnet 5.5
LMArena WebDev (LMArena)13941772
LMArena Elo14561473
Chess Puzzles (Epoch AI)10%59%
FrontierMath Tier 4 v2 (Epoch AI)2.4%80.5%
FrontierMath Tiers 1-3 v2 (Epoch AI)23.9%88.8%
GPQA Diamond83.4%95.6%
Mystery Game Puzzles (Epoch AI)17%65%
OTIS Mock AIME 2024-2025 (Epoch AI)77.8%100%
SAGE (Vals AI)36.1%51.8%
SimpleBench (SimpleBench)54.3%75.9%
SimpleQA Verified30.7%46.5%
Vending-Bench 2 (Andon Labs)3838.7411165.28
Humanity's Last Exam (no tools)17.7%56.9%
Humanity's Last Exam (with tools)33.6%64.5%

FAQ

Is Claude Sonnet 4.5 better than Claude Sonnet 5.5?

Claude Sonnet 5.5 scores 73.4 on the Respan Index (#6), ahead of Claude Sonnet 4.5 at 31.3 (#86). Claude Sonnet 5.5 leads on 14 of the 14 benchmarks both report.

Which is cheaper, Claude Sonnet 4.5 or Claude Sonnet 5.5?

Claude Sonnet 4.5 lists at $3 input and $15 output per 1M tokens; Claude Sonnet 5.5 at $2 and $10. Claude Sonnet 5.5 is 1.5x cheaper per token at list price.

Which has the larger context window?

Claude Sonnet 4.5 accepts 200K tokens and Claude Sonnet 5.5 1M tokens.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.