SShogo Bench

Qwen2.5-7B

Qwen2.5-7B from Alibaba (Qwen).

Shogo Agent Index

Unranked

Not ranked yet: 0 agentic benchmarks measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.

Agentic coding
6
estimated
Long-horizon autonomy
12
estimated
Knowledge-work agents
8
estimated
Computer use
0
estimated
Security & context learning
4
estimated

Price

No public API price.

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

Agentic benchmarks

BenchmarkScoreNormalisedIndex weightHarness / source
SWE-bench Verifiedest.8%predicted 20
Terminal-Bench 2.0est.8%predicted 0
DeepSWEest.6%predicted 1
FrontierSWEest.5%predicted 0
CursorBenchest.4%predicted 6
GSO-Benchest.4%predicted 0
METR Time Horizonsest.10%predicted 23
Vending-Bench 2est.6%predicted 0
PostTrainBenchest.4%predicted 4
APEX-Agentsest.8%predicted 1
GDPvalest.7%predicted 4
Remote Labor Indexest.5%predicted 0
DeepResearch Benchest.5%predicted 34
OSWorld 2.0est.6%predicted 0
OSWorldest.4%predicted 1
Cybenchest.5%predicted 2
CL-benchest.5%predicted 6

Other benchmarks

BenchmarkCategoryScoreSource
BALROGLong-horizon7.8%source
DTBenchKnowledge47.7%source
GPQA DiamondReasoning35.5%source
LMCAKnowledge6.4%source
MMLULegacy72.9%source
OTIS Mock AIMEMath2.5%source