Model · OpenAI

OpenAI Gpt Luna 5.6

1 appearance · meta-score 64.7/100 · last seen 2026-07-14 · scored under rubric v15

Pooled composite across 1 primary run: 61.1 — no interval, a single run proves nothing about spread provisional: n < 3, treat as unsettled

$26.22Total cost
83,035,988Tokens
82.2%Cache hit
1,568Requests
0.00%Failure rate
1.14Tools/response
3,166,753Tokens per $

Agent role scores

Per-agent-role scores for OpenAI Gpt Luna 5.6
Agent role	Score / 10
Sprint-Plan Agent	7.99
Scout Agent	7.82
Test Agent	7.40
Orchestrator	7.03
Code-Review Agent	6.67
Sprint-Review Agent	6.34
Develop Agent	6.28
Pseudocode Agent	6.24
Build Agent	5.84
Code Agent	5.58
Platform Agent	4.95

Run history (primary model)

Runs where OpenAI Gpt Luna 5.6 was the primary model, oldest first
Run	SoW	Date	Composite
Gpt Luna 5.6	circles	Jul 2026	61.1

Every run on Favur Evals is scored by the same deterministic engine, on the same Statements of Work. The benchmark is self-funded — no vendor sponsorship, credits, or grants.