Configuration comparison

GPT-5.6 Sol vs Qwen 3.8

A same-batch comparison of the highest-scoring measured configuration from each family, with route and missing-cost boundaries left visible.

Current batch

GPT-5.6 Sol / Max leads the selected batch by 19 points. Qwen 3.8 / Max completed 23m 39s sooner. Both configurations have reference-cost coverage.

Open the complete Radar Read the evaluation method

Selected evidence

Published: Aug 26, 2026, 11:00 PM UTC
Question pack: coding-fast-v4.10
Coverage: 2 measured configurations
Batch: snapshot-2026-08-26T23-00-00Z-r3

Same-batch evidence

Measure	GPT-5.6 Sol / Max	Qwen 3.8 / Max
Rank	#1	#10
Score	85/100	66/100
Elapsed	44m 58s	21m 19s
Reference cost	$1.82	$0.89
Route	Official login	Custom endpoint
Black-box Regression Audit	14/20	10/20
Retry Planner Counterexamples	20/20	17/20
CI Adversarial Audit	18/20	12/20
Transaction Regression Design	17/20	13/20
Cache Regression Test Design	16/20	14/20

The decision this comparison can support

Does the measured quality difference survive the route, elapsed-time, and cost-coverage trade-off for your work?

Start with the score and question profile. Use elapsed time only after both configurations clear your quality floor, and leave cost undecided when either side is not covered.