Benchmarks Public, governed benchmark suites generated from reference environments and verified against live runtime truth. benchmarks benchmarks benchmarks benchmarks/_index.md
GraphJin

Benchmarks

Benchmarks

Public, governed benchmark suites generated from reference environments and verified against live runtime truth.

GraphJin benchmarks are frozen, oracle-verified suites generated from reference environments. Models run through the same governed GraphJin interface, while publication gates preserve the suite fingerprint, reject stale scoring contracts, flag suspicious answer/method divergence, and record the exact GraphJin build used for every result.

The family grows as real reference environments are added. We list only benchmarks that exist and have a published, reproducible suite.

BenchmarkStatusModels on boardLast updated
DeepORG — The Organizational Agent BenchmarkLiveGemini 3.5 Flash-Lite, OpenAI GPT-5.4 mini2026-08-08
Docs
Start
Install Quick Start Demos First Query Saved Queries
Story
Vision Security Agentic
Core
Compiler Model Query Language Filters Ordering And Cursors Relationships Aggregations And Functions Mutations Subscriptions
Integrations
Multi-Database MongoDB OpenAPI Filesystem Tables And Uploads CodeSQL Apollo Federation
Agentic
Server-Side Agent Agent Evaluation MCP Catalog Graph Security Graph Source Mode Artifacts Overlay Declared Tasks Catalog Annotations Watches Watch Automation Mission Control Workflows MCP OAuth
Configure
How Configuration Works Sources Mode Database Config Auth And RBAC Caching And Redis Discovery Cache And Semantic Search Uploads And Filesystems OpenAPI Config Environment And Production
Reference
Database Support Config Reference Operators And Directives Troubleshooting Test-Backed Examples
Benchmarks
Methodology