GPT-6 Sol (max) vs MiniMax M3
Head-to-head on the Towards AI writing benchmark: same 10 real YouTube scripts, five runs each, scored blind by a three-family judge panel. GPT-6 Sol (max) leads overall, 88.0 to 82.7.
The verdict
GPT-6 Sol (max) wins this one, and the board is not close: #31 at 2165.6 Elo and 87.95 overall, against #71, 1672.0 Elo and 82.70 overall for MiniMax M3. The 95% Elo intervals do not overlap (2127.0 to 2200.7 against 1632.1 to 1702.1). Where GPT-6 Sol (max) pulls ahead: Slop Score (EQ-Bench + ours) (94.08 vs 85.83) and Substance, Accuracy & Value (88.87 vs 81.39). MiniMax M3 does not take a single metric off GPT-6 Sol (max) in this pairing. The trade is price: MiniMax M3 costs about $0.023 per article against $0.112 for GPT-6 Sol (max), so you pay for the higher rank. MiniMax M3 publishes open weights; GPT-6 Sol (max) does not. Default to GPT-6 Sol (max). Reach for MiniMax M3 when the lower price and open weights matter more to you than the headline rank.
Pick GPT-6 Sol (max) for the stronger board result, slop score, and substance.
Pick MiniMax M3 for the lower price and open weights.
Reading the Elo intervals. Each range estimates uncertainty around one model's Elo. We repeatedly sample from every model's recorded run scores on each task and recalculate the full ranking. Comparing these separate ranges does not test the Elo gap between two models. The tasks stay fixed, so the ranges do not measure how the ranking would change on new tasks.
Metric by metric
Blue bars: GPT-6 Sol (max). Orange bars: MiniMax M3. Same 0–100 scale; the bold bar wins that metric.
Everything else that differs
| GPT-6 Sol (max) | MiniMax M3 | |
|---|---|---|
| Overall / 100 | 88.0 | 82.7 |
| Writing Elo | 2166 | 1672 |
| Run-to-run spread (± overall std) | 1.330 | 1.980 |
| Cost per script (USD) | 0.112 | 0.023 |
| Avg latency (s) | 171.0 | 109.9 |
| Open weights | No | Yes |
Full scorecards: GPT-6 Sol (max) · MiniMax M3. How scoring works: methodology.
← All comparisons