Claude Opus 4.1 vs Claude Opus 4.5

Wins 1 of 7 benchmarks
Wins 2 of 7 benchmarks
Terminal-Bench Hard (version unstated) Terminal-Bench Hard (version unstated): Agents solve difficult software, system-administration and data-processing tasks in a terminal
Opus 4.134.3
Opus 4.540.9
AA-LCR (long context) 1.1 AA-LCR (long context) 1.1: Questions that need reading and connecting facts across very long documents (around 100K tokens)
Opus 4.176
Opus 4.570.7
SWE-bench Pro (private) SWE-bench Pro (private): Agents resolve long-horizon professional repository tasks; public and private splits are separate
Opus 4.117.8
Opus 4.523.4
MCP-Atlas Public MCPAtlas Public: MCPAtlas Public as reported in this launch evaluation; the post is the authority for the recorded task and grading scope
Opus 4.1—
Opus 4.5 · high73.4
MCP-Atlas · full 1000 MCP-Atlas · full 1000: Tool orchestration across MCP servers on the stated task split
Opus 4.1—
Opus 4.5 · high69.8
τ³-bench · banking · 1.0.1 τ-bench official · 3-banking-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools
Opus 4.1—
Opus 4.5 · high24.7
EBR Bench EBR Bench: Agents learn to play Earthborne Rangers across repeated playthroughs
Opus 4.17.9
Opus 4.5—

Specs

SpecOpus 4.1Opus 4.5
Input, per 1M$15$5
Output, per 1M$75$25
Speed, tok/s—51
Context200K200K
WeightsClosedClosed
ReleasedAug 2025Nov 2025
See images✓✓
Read PDFsPartialPartial
Hear audio——
Watch video——
Call tools✓✓
Force a tool callPartialPartial
Return strict JSON—✓
Search the web??
Run code??
Use a computer??

What breaks when you switch

Opus 4.1 → Opus 4.5
Nothing known.
Opus 4.5 → Opus 4.1
No strict JSON