SShogo Bench

Claude Haiku 5.5 vs Claude Opus 5.5

Claude Opus 5.5 leads Claude Haiku 5.5 by 13.4 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Haiku 5.5 wins 0 and Claude Opus 5.5 wins 21 of 21 shared benchmarks. Claude Haiku 5.5 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Haiku 5.5Claude Opus 5.5
Agentic coding6781
Long-horizon autonomy6276
Knowledge-work agents4658
Computer use4766
Security & context learning5562

Shared benchmarks

BenchmarkClaude Haiku 5.5Claude Opus 5.5
APEX-Agentsindex56.1%73.5%
Vending-Bench 2index$3,602$9,235
FrontierSWEindex43.8%62.3%
CursorBenchindex48.4%57.8%
ARC-AGI89.0%98.5%
ARC-AGI-261.3%93.3%
Chess Puzzles55.0%67.0%
CritPt22.6%31.7%
DTBench95.0%98.9%
FrontierCode46.4%54.6%
FrontierMath Tier 446.3%95.0%
FrontierMath Tiers 1-375.1%91.2%
GPQA Diamond89.6%94.1%
LMCA51.1%68.2%
MirrorCode78.9%95.6%
OTIS Mock AIME98.9%100.0%
ProofBench89.0%100.0%
SciCode55.0%66.9%
SimpleBench59.2%88.4%
SimpleQA Verified23.8%72.2%
WeirdML v311.7%31.2%

Specs and price

Claude Haiku 5.5Claude Opus 5.5
Agent Index57.5 (#18)70.9 (#3)
Agent $/M$0.050$1.69
Input / output $/M$0.10 / $0.50$4.00 / $20.00
Context1M1M
ReleasedOct 7, 2026Sep 22, 2026
WeightsClosedClosed
API hosts56

Don’t pick one. Route.

Shogo can run Claude Haiku 5.5 and Claude Opus 5.5 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo