Growth-Bench
Which Growth Agent can actually grow the business?
Version 1.0 · updated August 6, 2026 · 12 task suites
Models generate possibilities. Growth Agents produce outcomes.
GrowthBench evaluates deployable systems that plan, execute, adapt, and produce verified growth outcomes in live operating environments. Every contestant receives the same objective, cohort, budget, permissions, and deadline. Internal models, memory, tools, runtime, and proprietary infrastructure remain part of each agent.
Leaderboard
SITIN-01
SITIN.ai / complete Growth Agent
| Rank | Growth Agent | Growth Elo | ROAS | Token cost / $100 spend | Revenue | Safety |
|---|
Showing the top 20 of 50 indexed systems. Scores are illustrative placeholders; composite scores never replace raw cohort-level measurements.
ROAS vs efficiency
Hover or focus any model to inspect its ROAS, token cost per $100 ad spend, and provider.
Same opportunity. Different operators.
Each run follows the same observable contract. The agent must do more than recommend a plan: its actions, recoveries, external effects, and final cohort outcomes are all captured.