Grok 4.6 (high)
xAI · proprietary · writing benchmark
The short version
Grok 4.6 (high) sits at #18 of 134 on ToneBench, in the upper tier, with a writing Elo of 2144 and an overall score of 88.1 out of 100. We measured it by having it write all 10 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 xAI, this is the strongest writer we tested. It edges out the other 8 xAI 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 16th of 85. That ranking is against the stronger half of the board: closed models still set the pace on voice fidelity here.
One thing that stands out: it is unusually consistent. Across its 50 drafts the overall score only moves ± 1.4 points run to run, versus a board median of ± 3.6. What you get on the first try is close to what you get every try.
What it does best is Hook: it ranks 8th on that dimension at 90.0, well above the board average. Its softer spot is Visual Cues (86.4, 30th), which is the thing to watch if that metric matters most for your use.
It held steady across all 10 scripts, with its best run on the career guide / hiring analysis (89.5). Consistency like that is underrated, honestly, since it means fewer surprises when you actually put it to work.
On cost, it runs about $0.115/task. For that money it beats 7 models that cost noticeably more.
If you're tuning reasoning effort, we also tested Grok 4.6 at other settings. None of them beat this config; the closest is Grok 4.6 (Elo 2110), so the only question is how much quality you want to trade for cost or speed.
Skill profile
The shape is the story: the further a corner reaches, the stronger that dimension. Biggest edge over the board: Length Discipline (+20.8); smallest edge: Anti-Slop (+9.0).
Per-metric scores
Nine writing dimensions, each scored 0–100 and blended by the editorial weight shown. Rank is against all 134 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 10 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, Grok 4.6 (high) varies by ± 1.4 points between runs versus a board median of ± 3.6, so it is noticeably steadier than the typical model here.
Its steadiest is Tone & Voice (± 1.0 vs ± 2.2 board median), which you can rely on run after run.
| Metric | This model | Board median | Verdict |
|---|---|---|---|
| Tone & Voice ? | 88.8 ± 1.0 | ± 2.2 | steadier than most |
| Writing Craft ? | 88.0 ± 1.2 | ± 2.0 | steadier than most |
| Substance & Value ? | 88.0 ± 1.4 | ± 2.5 | steadier than most |
| Flow & Emotion ? | 86.3 ± 1.5 | ± 2.6 | typical |
| YouTube Structure ? | 87.1 ± 1.5 | ± 3.6 | steadier than most |
| Hook ? | 90.0 ± 1.3 | ± 2.9 | steadier than most |
| Length Discipline ? | 88.3 ± 11.5 | ± 8.9 | typical |
| Anti-Slop ? | 91.8 ± 1.4 | ± 2.5 | steadier than most |
| Visual Cues ? | 86.4 ± 2.9 | ± 4.8 | steadier than most |
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 88.1 is the consensus of three family-disjoint judges scoring the same 50 stored drafts blind with the identical rubric. The highest and lowest judge differ by 4.7 points on its overall. The three judges essentially agree on this model. How the panel works: methodology.
How we ran Grok 4.6 (high)
- Accessed viaxAI API · model page ↗
- Exact model / route id
grok-4.6atreasoning_effort: high - Generations5 independent runs × 10 scripts = 50 drafts, stored immutably after generation
- Run dates2026-08-13
- 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 Grok 4.6 (high) at writing?
On ToneBench it ranks #18 of 134 with a writing Elo of 2144 and an overall score of 88.1/100. That score comes from writing our 10 real YouTube scripts five times each and scoring every draft blind against our own finished versions across nine writing dimensions.
Is Grok 4.6 (high) the best xAI model for writing?
Yes. Among the xAI configs we tested, Grok 4.6 (high) is the strongest writer on the board.
Is Grok 4.6 (high) good value for the money?
It costs about $0.115/task. Nothing meaningfully cheaper outscores it, which puts it on the value side of the board.
What are Grok 4.6 (high)'s strengths and weaknesses?
Its strongest dimension is Hook (8th on the board, 90.0). Its weakest is Visual Cues (30th on the board, 86.4). The full nine-metric breakdown is on this page.
Does Grok 4.6 (high) write some formats better than others?
It was fairly even across all 10 scripts, topping out on the career guide / hiring analysis (89.5).
How consistent is Grok 4.6 (high) between runs?
We run every script 5 times. Grok 4.6 (high)'s overall score varies by about ±1.4 points between runs, versus a board median of ±3.6. That makes it one of the steadier models we test. The full per-metric variability table is on this page.
How was Grok 4.6 (high) evaluated?
Via xAI API using the exact model/route id grok-4.6 at reasoning_effort high, run on 2026-08-13. 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 Grok 4.6 (high) open source?
No, it is a proprietary (closed-weights) model. Among proprietary models it ranks 16th of 85 for writing.