DeepSeek V4 Flash 0731 vs GLM-5
Should you write with DeepSeek V4 Flash 0731 or GLM-5? 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. DeepSeek V4 Flash 0731 leads overall, 84.4 to 82.8.
The verdict
The cheaper model wins this one. DeepSeek V4 Flash 0731 ranks #59 to GLM-5's #66, and GLM-5 is behind on every metric. At $0.005 per script, DeepSeek is also about ten times cheaper than GLM-5's $0.050. The clearest gap is Length Adherence, where GLM-5 lands more than seven points lower, and that's the kind of miss that turns a tight script into an editing job. Both are open weights and they run at about the same speed, so nothing else pulls you toward GLM-5. If you like the GLM family, look at GLM-5.3 instead, which sits far higher on the board.
Pick DeepSeek V4 Flash 0731 for cheap open-weights scripts that stay closer to the target length.
Pick GLM-5 only if it's the model you already host; otherwise GLM-5.3 is the better GLM.
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 DeepSeek V4 Flash 0731, orange is GLM-5, 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, Length Adherence does the most to put DeepSeek V4 Flash 0731 ahead: it leads there by 7.6 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. DeepSeek V4 Flash 0731 is ahead and it's also the cheaper one: GLM-5 costs about 10x as much per script.
| DeepSeek V4 Flash 0731 | GLM-5 | |
|---|---|---|
| Overall / 100 | 84.4 | 82.8 |
| Writing Elo | 1866 | 1692 |
| Score spread across drafts (± overall std) | 3.360 | 2.400 |
| Cost per script (USD) | 0.005 | 0.050 |
| Avg time per script (min) | 3.1 | 3.1 |
| Open weights | Yes | 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: DeepSeek V4 Flash 0731 · GLM-5. The whole scoring pipeline is on the methodology page.
← All comparisons