Qwen3.7 Max (high) vs DeepSeek V4 Pro (xhigh)
Should you write with Qwen3.7 Max (high) or DeepSeek V4 Pro (xhigh)? On this page the only thing that changes is the model. Both wrote their own version of the scripts behind 10 of my What's AI videos, five drafts per script, and three judges from three model families scored every draft blind, with my finished script as the reference. Qwen3.7 Max (high) leads overall, 80.8 to 78.9.
The verdict
I'd take Qwen3.7 Max (high) over DeepSeek V4 Pro (xhigh), though neither is where I'd start for scripts. Qwen ranks #89, DeepSeek #97, about 108 Elo apart, and here the ranges are worth a look: they barely clear each other, so the order is likely, just not by much. Qwen's edge is structure: visual cues by about eight points and YouTube Best Practices by about six, plus less slop. DeepSeek keeps a slight lead on voice and writing craft, under a point each. It's also cheaper, about $0.013 a script against $0.055, and it's open weights where Qwen isn't. So the trade is real, just lopsided. If you'll restructure the draft yourself and want open weights at a low price, DeepSeek is defensible.
Pick Qwen3.7 Max (high) when you want better cues and YouTube structure from a budget model.
Pick DeepSeek V4 Pro (xhigh) when open weights and a lower price matter and you'll rebuild the structure yourself.
How to read these numbers
Overall vs Elo. Overall is the weighted rubric score. Elo compares each model's five-draft average with every other model's, script by script, and counts gaps inside run-to-run noise as draws, so it rewards winning often and can disagree with overall. My own scripts score 90.9 on the same rubric, not a literal 100. Right now 2 of 167 ranked configurations reach it. A model above that line writes at that level on this rubric. That's not the same as beating my script line for line.
Reading the Elo intervals. Each 95% range comes from resampling every model's recorded run scores on each task and re-ranking. Comparing these separate ranges does not test the Elo gap between two models: far apart is strong evidence, and overlap isn't a tie. The tasks stay fixed, so the ranges say nothing about new tasks (the long version).
Metric by metric: where the gap comes from
Blue is Qwen3.7 Max (high), orange is DeepSeek V4 Pro (xhigh), and the solid bar takes the metric. The weight under each name is how much it counts toward the overall. Weighted by how much each metric counts, YouTube Best Practices does the most to put Qwen3.7 Max (high) ahead: it leads there by 6.2 points. If you care about one thing, say voice or length, go straight to that row.
Cost, speed, and consistency
Quality is half the decision. The other half is what each script costs you, how long you wait, and how often you get a bad draft. The lead has a price: Qwen3.7 Max (high) costs about 4.2x as much per script as DeepSeek V4 Pro (xhigh).
| Qwen3.7 Max (high) | DeepSeek V4 Pro (xhigh) | |
|---|---|---|
| Overall / 100 | 80.8 | 78.9 |
| Writing Elo | 1525 | 1417 |
| Score spread across drafts (± overall std) | 2.190 | 3.840 |
| Cost per script (USD) | 0.055 | 0.013 |
| Avg time per script (min) | 1.7 | 1.8 |
| Open weights | No | Yes |
Green wins the row, and for spread, cost and time, lower wins. Cost is one script at recorded list prices, uncached and without judging, so your real bill will differ. Spread is how much the overall score moves across a model's drafts, and time is the average time to get one accepted script, retries included.
Want the script-by-script detail? Full scorecards: Qwen3.7 Max (high) · DeepSeek V4 Pro (xhigh). The whole scoring pipeline is on the methodology page.
← All comparisons