SShogo Bench

GLM-4.6

Late GLM-4 workhorse for coding agents, reasoning, and structured tasks

Shogo Agent Index

Unranked

Not ranked yet: 1 agentic benchmark measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.

Agentic coding
26
23% measured
Long-horizon autonomy
35
estimated
Knowledge-work agents
22
estimated
Computer use
14
estimated
Security & context learning
26
estimated

Price

Input
$0.60/M
Output
$2.20/M
Cache read
$0.11/M
3:1 blend
$1.00/M
Agent-effective
$0.28/M

Cheapest host for agents: DeepInfra at $0.20/M.

Agent-effective = 20 input tokens per output token, 85% cache hits. Why.

Agentic benchmarks

BenchmarkScoreNormalisedIndex weightHarness / source
SWE-bench Verifiedest.8%predicted 60
Terminal-Bench 2.024.5%8%Terminus 2
DeepSWEest.6%predicted 18
FrontierSWEest.5%predicted 2
CursorBenchest.4%predicted 23
GSO-Benchest.4%predicted 5
METR Time Horizonsest.10%predicted 59
Vending-Bench 2est.6%predicted 7
PostTrainBenchest.4%predicted 15
APEX-Agentsest.8%predicted 19
GDPvalest.7%predicted 25
Remote Labor Indexest.5%predicted 1
DeepResearch Benchest.5%predicted 43
OSWorld 2.0est.6%predicted 1
OSWorldest.4%predicted 33
Cybenchest.5%predicted 37
CL-benchest.5%predicted 15

Other benchmarks

BenchmarkCategoryScoreSource
CritPtReasoning1.1%source
SciCodeCoding38.4%source
WebDev ArenaCoding1341source

API hosts

Live prices and uptime for every host serving GLM-4.6 via OpenRouter.

HostInputOutputCache readAgent $/MUptime 24hQuant
DeepInfra$0.40$1.60$0.080$0.20—fp4
Venice$0.43$1.75$0.080$0.2199.4%fp4
DeepInfra$0.50$2.00$0.10$0.2599.9%fp4
Novita$0.55$2.20$0.11$0.2799.2%bf16
Z.AI$0.60$2.20$0.11$0.2899.5%fp4