SShogo Bench

Claude Haiku 5.5 vs GLM-5.3

Claude Haiku 5.5 leads GLM-5.3 by 2.3 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Haiku 5.5 wins 10 and GLM-5.3 wins 7 of 17 shared benchmarks. Claude Haiku 5.5 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Haiku 5.5GLM-5.3
Agentic coding6762
Long-horizon autonomy6268
Knowledge-work agents4643
Computer use4740
Security & context learning5551

Shared benchmarks

BenchmarkClaude Haiku 5.5GLM-5.3
APEX-Agentsindex56.1%56.6%
Vending-Bench 2index$3,602$8,164
FrontierSWEindex43.8%30.2%
CursorBenchindex48.4%42.6%
Chess Puzzles55.0%37.0%
CritPt22.6%19.1%
DTBench95.0%87.7%
FrontierCode46.4%40.1%
FrontierMath Tier 446.3%29.3%
FrontierMath Tiers 1-375.1%68.8%
GPQA Diamond89.6%90.9%
LMCA51.1%55.5%
OTIS Mock AIME98.9%91.1%
ProofBench89.0%49.0%
SciCode55.0%59.0%
SimpleBench59.2%66.2%
SimpleQA Verified23.8%41.0%

Specs and price

Claude Haiku 5.5GLM-5.3
Agent Index57.5 (#18)55.2 (#25)
Agent $/M$0.050$0.61
Input / output $/M$0.10 / $0.50$1.20 / $4.18
Context1M1M
ReleasedOct 7, 2026Aug 14, 2026
WeightsClosedOpen
API hosts549

Don’t pick one. Route.

Shogo can run Claude Haiku 5.5 and GLM-5.3 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo