MiniMax M3
MiniMax · open weights · writing benchmark
The short version
MiniMax M3 sits at #42 of 130 on ToneBench, in the middle of the pack, with a writing Elo of 1790 and an overall score of 82.4 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 MiniMax, this is the strongest writer we tested. It edges out the other 2 MiniMax configs on the board, so if you're staying in this family for voice work, this is the one to reach for.
Among open-weights models, it comes in 5th of 47. That is a genuinely strong showing for an open-weights model on a voice-and-tone task, which is usually where the closed frontier still pulls ahead.
What it does best is Length Discipline: it ranks 38th on that dimension at 82.6, well above the board average. Its softer spot is YouTube Structure (77.7, 64th), 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 (87.3) and its weakest was the personal roadmap / opinion (75.5), a spread of about 12 points. Worth knowing if your writing skews toward one of those formats.
On cost, it runs about $0.019/task (estimated). DeepSeek V4 Flash 0731 scores higher for less money, at $0.005/task, so on pure value this config is not the frontier. Still, it beats 35 models that cost noticeably more.
Skill profile
The shape is the story: the further a corner reaches, the stronger that dimension. Biggest edge over the board: Length Discipline (+15.5); smallest edge: Substance & Value (+3.2).
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, MiniMax M3 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 most volatile dimension is Substance & Value (± 4.0 vs a board median of ± 2.6): two runs of the same brief can land visibly different substance & value scores.
| Metric | This model | Board median | Verdict |
|---|---|---|---|
| Tone & Voice ? | 83.8 ± 2.2 | ± 2.2 | typical |
| Writing Craft ? | 83.9 ± 2.3 | ± 2.0 | typical |
| Substance & Value ? | 81.8 ± 4.0 | ± 2.6 | swingier than most |
| Flow & Emotion ? | 79.2 ± 4.1 | ± 2.6 | swingier than most |
| YouTube Structure ? | 77.7 ± 6.1 | ± 3.6 | swingier than most |
| Hook ? | 86.1 ± 2.3 | ± 3.0 | typical |
| Length Discipline ? | 82.6 ± 12.8 | ± 8.6 | typical |
| Anti-Slop ? | 87.3 ± 2.1 | ± 2.6 | typical |
| Visual Cues ? | 83.0 ± 5.5 | ± 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 82.4 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 7.7 points on its overall. A typical amount of judge disagreement for this board. How the panel works: methodology.
How we ran MiniMax M3
- Accessed viaOpenRouter · model page ↗
- Exact model / route id
minimax/minimax-m3(provider default reasoning) - Audited route overrideOpenRouter ·
minimax/minimax-m3forautomated_video_workflow(Task-7 exact OpenRouter endpoint pin novita/fp8) - Generations5 independent runs × 9 scripts = 45 drafts, stored immutably after generation
- Run dates2026-07-29 → 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 MiniMax M3 at writing?
On ToneBench it ranks #42 of 130 with a writing Elo of 1790 and an overall score of 82.4/100. 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 MiniMax M3 the best MiniMax model for writing?
Yes. Among the MiniMax configs we tested, MiniMax M3 is the strongest writer on the board.
Is MiniMax M3 good value for the money?
It costs about $0.019/task (estimated). DeepSeek V4 Flash 0731 scores higher for less, so it is not the value pick.
What are MiniMax M3's strengths and weaknesses?
Its strongest dimension is Length Discipline (38th on the board, 82.6). Its weakest is YouTube Structure (64th on the board, 77.7). The full nine-metric breakdown is on this page.
Does MiniMax M3 write some formats better than others?
Yes. Its best of our 9 scripts was the founder announcement / personal origin story (87.3) and its weakest was the personal roadmap / opinion (75.5).
How consistent is MiniMax M3 between runs?
We run every script 5 times. MiniMax M3'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 MiniMax M3 evaluated?
Via OpenRouter using the exact model/route id minimax/minimax-m3, run on 2026-07-29. 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 MiniMax M3 open source?
Yes, it is an open-weights model. Among open-weights models it ranks 5th of 47 for writing.