Where do these coding results come from?

Both coding agents worked the same 24 merged changes from the private Stet-AI/stet repository, each in the real repo as it was before the change. The repo’s tests and judges graded every patch against the merged change; no public benchmark scores are used.

Are reasoning settings combined into one model score?

No. Each effort pairing keeps its own recorded quality call, head-to-head record, cost and agent time. Incompatible scores are never averaged.

GPT 5.6 Luna vs Grok 4.7

GPT 5.6 Luna cheaper in both pairings. Grok 4.7 leads on quality in 1 of 2 pairings; the rest show no clear difference. 2 pairings; 24 merged changes in Stet-AI/stet. Compare the individual calls below to choose which setup to investigate.

What this comparison is

1 real repositoryStet-AI/stet, private; not public benchmark scores
24 merged changeseach run end to end by the coding agent from the repo as it was before the change
Graded against the merged changeby the repo’s own tests and by judges comparing each patch with it
Quality, cost and timea smaller sample than public benchmarks, but real agent work on a real codebase

Compare coding models Low vs high vs xhigh/max reasoning for coding agents

Recorded setup pairings

Each row is a separate recorded comparison. Scores are not pooled or averaged across effort settings. W/D/L means wins, draws and losses; undecided judgments remain undecided.

Stet score: A setup’s average chance of beating another setup on a task, from 0 to 100; 50 is even with the field.

Setup pairing	Recorded quality call and record	Quality evidence	Cost per task	Agent time per task
GPT 5.6 Luna max vs Grok 4.7 highGPT 5.6 Luna max · max reasoningGrok 4.7 high · high reasoning24/24 tasks observed by both setups.	No clear quality difference: Grok 4.7 high won 3, GPT 5.6 Luna max won 2, 18 tied, 1 undecided; GPT 5.6 Luna max leads on cost: 6.4× cheaper; Grok 4.7 high leads on agent time: 1.5× faster.Head to head, Grok 4.7 high won 3, tied 18 and lost 2 against GPT 5.6 Luna max, with 1 undecided.	No clear difference 2–3 head to head, 18 tiedGPT 5.6 Luna max: 49.7 (95% range 44.0–55.6)Grok 4.7 high: 52.9 (95% range 46.9–58.7)	GPT 5.6 Luna max: $0.90 Grok 4.7 high: $5.76	GPT 5.6 Luna max: 28.5 min Grok 4.7 high: 18.9 min
GPT 5.6 Luna low vs Grok 4.7 highGPT 5.6 Luna low · codex · low reasoningGrok 4.7 high · high reasoning24/24 tasks observed by both setups.	Grok 4.7 high leads on quality, 4–1 head to head with 18 tied and 1 undecided (directional at this sample size); GPT 5.6 Luna low leads on cost and agent time: 113.1× cheaper and 2.8× faster.Head to head, Grok 4.7 high won 4, tied 18 and lost 1 against GPT 5.6 Luna low, with 1 undecided.	Directional lead at this sample size 4–1 head to head, 18 tiedGPT 5.6 Luna low: 44.6 (95% range 38.3–50.3)Grok 4.7 high: 52.9 (95% range 46.9–58.7)	GPT 5.6 Luna low: $0.05 Grok 4.7 high: $5.76	GPT 5.6 Luna low: 6.7 min Grok 4.7 high: 18.9 min

Setup pairing

Recorded quality call and record

Quality evidence

Cost per task

Agent time per task

GPT 5.6 Luna max vs Grok 4.7 highGPT 5.6 Luna max · max reasoningGrok 4.7 high · high reasoning24/24 tasks observed by both setups.

No clear quality difference: Grok 4.7 high won 3, GPT 5.6 Luna max won 2, 18 tied, 1 undecided; GPT 5.6 Luna max leads on cost: 6.4× cheaper; Grok 4.7 high leads on agent time: 1.5× faster.Head to head, Grok 4.7 high won 3, tied 18 and lost 2 against GPT 5.6 Luna max, with 1 undecided.

No clear difference
2–3 head to head, 18 tiedGPT 5.6 Luna max: 49.7 (95% range 44.0–55.6)Grok 4.7 high: 52.9 (95% range 46.9–58.7)

GPT 5.6 Luna max: $0.90

Grok 4.7 high: $5.76

GPT 5.6 Luna max: 28.5 min

Grok 4.7 high: 18.9 min

GPT 5.6 Luna low vs Grok 4.7 highGPT 5.6 Luna low · codex · low reasoningGrok 4.7 high · high reasoning24/24 tasks observed by both setups.

Grok 4.7 high leads on quality, 4–1 head to head with 18 tied and 1 undecided (directional at this sample size); GPT 5.6 Luna low leads on cost and agent time: 113.1× cheaper and 2.8× faster.Head to head, Grok 4.7 high won 4, tied 18 and lost 1 against GPT 5.6 Luna low, with 1 undecided.

Directional lead at this sample size
4–1 head to head, 18 tiedGPT 5.6 Luna low: 44.6 (95% range 38.3–50.3)Grok 4.7 high: 52.9 (95% range 46.9–58.7)

GPT 5.6 Luna low: $0.05

Grok 4.7 high: $5.76

GPT 5.6 Luna low: 6.7 min

Grok 4.7 high: 18.9 min

Questions about these results

Which is better for coding, GPT 5.6 Luna or Grok 4.7?: GPT 5.6 Luna cheaper in both pairings. Grok 4.7 leads on quality in 1 of 2 pairings; the rest show no clear difference. 2 pairings; 24 merged changes in Stet-AI/stet. Compare the individual calls below to choose which setup to investigate.
Where do these coding results come from?: Both coding agents worked the same 24 merged changes from the private Stet-AI/stet repository, each in the real repo as it was before the change. The repo’s tests and judges graded every patch against the merged change; no public benchmark scores are used.
Are reasoning settings combined into one model score?: No. Each effort pairing keeps its own recorded quality call, head-to-head record, cost and agent time. Incompatible scores are never averaged.