Skip to main content

Claude Sonnet 4 vs Claude Sonnet 4.5

SpecClaude Sonnet 4Claude Sonnet 4.5
Respan Index21.0 (#122)31.3 (#85)
Blended price per 1M-$6
Context window200K200K
Input / output per 1M-$3 / $15
Max output64K64K
Inputtext, visiontext, vision
Released2025-05-222025-09-29

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 4Claude Sonnet 4.5
Coding8.3 Claude Sonnet 413.0 Claude Sonnet 4.5
Agents17.3 Claude Sonnet 441.5 Claude Sonnet 4.5
Reasoning23.7 Claude Sonnet 427.5 Claude Sonnet 4.5
Knowledge & language17.7* Claude Sonnet 431.2 Claude Sonnet 4.5
Human preference46.1 Claude Sonnet 451.0 Claude Sonnet 4.5

* Estimated: no published scores in that category. How the Respan Index works

Benchmarks both report

Claude Sonnet 4.5 leads on 10 of the 10 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
LMArena Elo14021456
GPQA Diamond79.2%83.4%
MATH Level 5 (Epoch AI)84.4%97.7%
OTIS Mock AIME 2024-2025 (Epoch AI)71.1%77.8%
SAGE (Vals AI)35%36.1%
SimpleBench (SimpleBench)45.5%54.3%
WeirdML (Håvard Tveit Ihle)46.1%47.7%
OSWorld-Verified (XLANG)41.4%62.9%
GSO Opt@1 (GSO)4.9%12.7%
SWE-bench Verified72.7%77.2%

FAQ

Is Claude Sonnet 4 better than Claude Sonnet 4.5?

Claude Sonnet 4.5 scores 31.3 on the Respan Index (#85), ahead of Claude Sonnet 4 at 21.0 (#122). Claude Sonnet 4.5 leads on 10 of the 10 benchmarks both report.

Which has the larger context window?

Both accept 200K tokens of context.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.