Qwen3.8 Max
Alibaba (Qwen) · proprietary · writing benchmark
The short version
Qwen3.8 Max sits at #38 of 130 on ToneBench, in the middle of the pack, with a writing Elo of 1866 and an overall score of 83.4 out of 100. We measured it by having it write all 9 of our real YouTube scripts, five times each, then scoring every draft blind against our own finished versions. Here is how it shook out.
Inside Alibaba (Qwen), this is the strongest writer we tested. It edges out the other 9 Alibaba (Qwen) configs on the board, so if you're staying in this family for voice work, this is the one to reach for.
Among proprietary models, it comes in 35th of 83. That ranking is against the stronger half of the board: closed models still set the pace on voice fidelity here.
What it does best is Length Discipline: it ranks 5th on that dimension at 92.9, well above the board average. Its softer spot is Hook (83.5, 69th), which is the thing to watch if that metric matters most for your use.
It was uneven across the 9 scripts. Its best run was the founder announcement / personal origin story (86.6) and its weakest was the personal technical walkthrough / agentic workflow case study (73.7), a spread of about 13 points. Worth knowing if your writing skews toward one of those formats.
On cost, it runs about $0.15/task. DeepSeek V4 Flash 0731 scores higher for less money, at $0.005/task, so on pure value this config is not the frontier.
Skill profile
The shape is the story: the further a corner reaches, the stronger that dimension. Biggest edge over the board: Length Discipline (+25.7); smallest edge: Hook (+3.3).
Per-metric scores
Nine writing dimensions, each scored 0–100 and blended by the editorial weight shown. Rank is against all 130 current-ranked models. The thick bar is this model; the thin lines above and below are the current field's best model (darker) and average (lighter) on the same scale. The small ± number is its run-to-run variation.
Per-article scores
The same 9 real scripts every current-ranked model writes, scored individually. Different formats stress different skills.
Consistency: run-to-run variability
Every cell on this page is the mean of 5 independent runs per script; the ± numbers are the run-to-run standard deviation. Overall, Qwen3.8 Max varies by ± 5.6 points between runs versus a board median of ± 3.8, so it is swingier than the typical model here, worth knowing if you need repeatable output.
Its steadiest is Anti-Slop (± 1.3 vs ± 2.6 board median), which you can rely on run after run.
| Metric | This model | Board median | Verdict |
|---|---|---|---|
| Tone & Voice ? | 81.9 ± 2.5 | ± 2.2 | typical |
| Writing Craft ? | 83.4 ± 2.1 | ± 2.0 | typical |
| Substance & Value ? | 84.4 ± 3.7 | ± 2.6 | typical |
| Flow & Emotion ? | 78.1 ± 3.5 | ± 2.6 | typical |
| YouTube Structure ? | 80.4 ± 4.9 | ± 3.6 | typical |
| Hook ? | 83.5 ± 2.7 | ± 3.0 | typical |
| Length Discipline ? | 92.9 ± 7.8 | ± 8.6 | typical |
| Anti-Slop ? | 92.0 ± 1.3 | ± 2.6 | steadier than most |
| Visual Cues ? | 85.2 ± 4.5 | ± 4.7 | typical |
Measured, not modeled
Numbers we log on every run and rarely talk about. None of these affect the writing scores; cost and latency are informational.
What each judge scored it
The published overall of 83.4 is the consensus of three family-disjoint judges scoring the same 45 stored drafts blind with the identical rubric. The highest and lowest judge differ by 10.5 points on its overall. A typical amount of judge disagreement for this board. How the panel works: methodology.
How we ran Qwen3.8 Max
- Accessed viaOpenRouter · model page ↗
- Exact model / route id
qwen/qwen3.8-max(provider default reasoning) - Generations5 independent runs × 9 scripts = 45 drafts, stored immutably after generation
- Run dates2026-08-05 → 2026-08-11
- Judges3-judge panel: Claude Opus 5, high effort (Anthropic) + DeepSeek V4 Flash (OpenRouter) + GPT-5.6 Sol, medium reasoning (Codex CLI): every draft scored independently by all three with the identical rubric, blind to model identity, evidence quotes required before each metric score; published score = per-metric consensus mean, verified with a Bradley-Terry refit (methodology)
- Samplingprovider-default sampling, no fine-tuning, no system-prompt changes; every current-ranked model gets the identical brief, style contract and research packet
- What stays privatethe exact rubric and any non-public reference material; source-publication timing is disclosed for exceptions (why)
Frequently asked questions
How good is Qwen3.8 Max at writing?
On ToneBench it ranks #38 of 130 with a writing Elo of 1866 and an overall score of 83.4/100. That score comes from writing our 9 real YouTube scripts five times each and scoring every draft blind against our own finished versions across nine writing dimensions.
Is Qwen3.8 Max the best Alibaba (Qwen) model for writing?
Yes. Among the Alibaba (Qwen) configs we tested, Qwen3.8 Max is the strongest writer on the board.
Is Qwen3.8 Max good value for the money?
It costs about $0.15/task. DeepSeek V4 Flash 0731 scores higher for less, so it is not the value pick.
What are Qwen3.8 Max's strengths and weaknesses?
Its strongest dimension is Length Discipline (5th on the board, 92.9). Its weakest is Hook (69th on the board, 83.5). The full nine-metric breakdown is on this page.
Does Qwen3.8 Max write some formats better than others?
Yes. Its best of our 9 scripts was the founder announcement / personal origin story (86.6) and its weakest was the personal technical walkthrough / agentic workflow case study (73.7).
How consistent is Qwen3.8 Max between runs?
We run every script 5 times. Qwen3.8 Max's overall score varies by about ±5.6 points between runs, versus a board median of ±3.8. That is swingier than typical, so expect more draft-to-draft variation. The full per-metric variability table is on this page.
How was Qwen3.8 Max evaluated?
Via OpenRouter using the exact model/route id qwen/qwen3.8-max, run on 2026-08-05. 5 runs per script, provider-default sampling, no fine-tuning; every draft scored blind with a fixed rubric by a three-family judge panel. Full details in the 'How we ran it' section and the methodology.
Is Qwen3.8 Max open source?
No, it is a proprietary (closed-weights) model. Among proprietary models it ranks 35th of 83 for writing.