Towards AITowards AIToneBench

Model Scorecard

DeepSeek V4.1 Flash (high)

DeepSeek · open weights

DeepSeek V4.1 Flash ranks #17 of 87 models on the November 2026 ToneBench board through its best configuration, DeepSeek V4.1 Flash (max). This setting, DeepSeek V4.1 Flash (high), is #48 of 151 configurations, with a Writing Score of 80.3/100 and a Writing Elo of 1882. One script costs $0.012 at API list price.

Last updated , from board data generated at 17:52 UTC.

Rank, scores, the nine metrics, the 10 reference scripts, cost and setup for one configuration (a model at one reasoning setting). Key Results sums it up; the section links below it jump to the detail.

Model Rank ?
#17 of 87 modelsthrough DeepSeek V4.1 Flash (max); this setting is #48 of 151 configurations
Writing Score ?
80.3 out of 100
Writing Elo ?
1882 95% range 1842 to 1921
Cost per Script
$0.012 at API list price

Key Results

Louis's PickI'd stay with this one: DeepSeek V4.1 Flash (max) is the strongest DeepSeek V4.1 Flash setting, but their Elo ranges overlap, and it costs under a cent more per script.

More Context Behind These Results

Writing Elo comes from comparing its average on each script with every other configuration's average on that same script; a gap smaller than the draft-to-draft variation counts as a draw (what the ranges can and cannot show).

On Length Adherence, its best-ranked metric, its drafts land near the target length, so they need less trimming or padding. Substance, Accuracy & Value ranks in the bottom half (89th of 151 configurations), so check the facts and the depth before anything goes out.

Its drafts vary little, so the first draft is usually close to the next one.

It outranks 67 configurations that cost at least 50% more.

Among open-weight models, DeepSeek V4.1 Flash is 5th of 34. This setting is 9th of 43 open-weight configurations. The best open-weight configuration is GLM-5.3 (max), at configuration #22, and every configuration above it is proprietary.

Lab
DeepSeek 2nd of 5 modelsthis setting 4th of 11 configurations
Weights
Open 5th of 34 modelsthis setting 9th of 43 configurations
Consistency
± 1.5 steadier than most
Average Tokens
30.1k input + output
Time per Script
1.1 min average, retries included
Highest-Ranked in This LabDeepSeek V4.1 Flash 1017 (max) ranks highest of the DeepSeek configurations by Writing Elo, with a Writing Score +1.3 against this configuration, for under a cent more per script.

All Settings of DeepSeek V4.1 Flash

DeepSeek V4.1 Flash is also on the board at other effort settings, and DeepSeek V4.1 Flash (max) ranks higher, though their Elo ranges overlap. The thinking-level page for DeepSeek V4.1 Flash charts how quality, cost and time change with the setting.

SettingConfiguration RankWriting ScoreWriting EloCost / ScriptTime / Script
max#3681.01956$0.0172.4 min
high this page#4880.31882$0.0121.1 min
low#6079.31784$0.00770.7 min
no thinking#8276.81562$0.00370.3 min

Skill Profile Across the Nine Metrics

Each corner is one metric on a 0 to 100 scale; the further out, the higher the score. A corner outside the dashed line beats the board median. Hover or tap a point for its exact score, and a metric name for what it rewards.

50Tone?Craft?Substance?Flow?YouTube?Hook?Length?Slop?Cues?
DeepSeek V4.1 Flash (high)Board medianBoard best per metricMax possible (100)

Scores by Metric

Nine metrics, each scored 0 to 100 and weighted into the Writing Score; the weight sits under each name. Rank is out of all 151 ranked configurations.

Score colors:Strong: 82 and upMiddle: 65 to 81.9Weak: under 65
This modelBoard best (upper line)Board median (lower line)±: spread between drafts?: what the metric rewards
Tone & Voice Match ?19% weight · +3.7 vs median
board best 82.8 · Claude Opus 5.5 (max effort)77.8± 1.4 · 40thboard median 74.0
Writing Craft & Clarity ?13% weight · +3.2 vs median
board best 83.2 · Claude Opus 5.5 (xhigh)78.2± 1.4 · 35thboard median 75.0
Substance, Accuracy & Value ?15% weight · -1.7 vs median
board best 88.4 · Claude Opus 5.5 (xhigh)80.2± 2.9 · 89thboard median 81.8
Continuity & Emotion ?14% weight · +2.6 vs median
board best 85.7 · Claude Opus 5.5 (max effort)80.1± 1.3 · 43rdboard median 77.5
YouTube Best Practices ?12% weight · +3.2 vs median
board best 84.5 · Claude Fable 5.2 (max effort)79.6± 2.5 · 40thboard median 76.4
Hook Strength ?10% weight · +0.4 vs median
board best 84.0 · Claude Fable 5.2 (xhigh)80.0± 2.9 · 70thboard median 79.6
Length Adherence ?8% weight · +8.4 vs median
board best 99.0 · Claude Fable 5.2 (xhigh)93.1± 4.3 · 34thboard median 84.8
Slop Score ?5% weight · +2.0 vs median
board best 89.0 · GPT-6 Sol (medium)83.3± 1.8 · 58thboard median 81.3
Visual Cue Quality ?4% weight · -0.1 vs median
board best 85.9 · Claude Opus 5.5 (max effort)74.6± 4.4 · 78thboard median 74.8

Scores by Reference Script

Every ranked configuration writes five drafts for each of the same 10 briefs, and each score is the mean of those drafts. The rank and board median compare it with every configuration on that script, which separates a weak result from a hard script. The result on the format closest to your own work is the most relevant.

Script 1
opinion / warning explainer
79.2
out of 100
60th of 151 configurations on this script · board median 77.2
Script 2
news-analysis / skeptical explainer
80.5
out of 100
53rd of 151 configurations on this script · board median 78.4
Script 3
personal roadmap / opinion
80.0
out of 100
48th of 151 configurations on this script · board median 77.2
Script 4
short explainer
79.9
out of 100
39th of 151 configurations on this script · board median 75.7
Script 5
news-analysis / opinion explainer
80.5
out of 100
48th of 151 configurations on this script · board median 78.0
Script 6
founder announcement / personal origin story
79.2
out of 100
76th of 151 configurations on this script · board median 79.2
Script 7
personal technical walkthrough / agentic workflow case study
81.8
out of 100
33rd of 151 configurations on this script · board median 77.9
Script 8
product release announcement / personal observation
82.0
out of 100
28th of 151 configurations on this script · board median 79.0
Script 9
career guide / hiring analysis
81.2
out of 100
60th of 151 configurations on this script · board median 79.3
Script 10
engineering process walkthrough / presentation adaptation
79.2
out of 100
40th of 151 configurations on this script · board median 75.0

Consistency Between Drafts

Every score on this page is the mean of 5 drafts per script. The ± is the standard deviation between those drafts: the larger it is, the more two drafts of the same script can differ.

Compared with the typical model, it varies most on Visual Cue Quality: ± 4.4 between drafts against ± 2.9, so two drafts of the same script can score quite differently there.

Consistency by Metric
MetricThis modelBoard medianVerdict
Tone & Voice Match ?77.8 ± 1.4± 1.9typical
Writing Craft & Clarity ?78.2 ± 1.4± 1.5typical
Substance, Accuracy & Value ?80.2 ± 2.9± 2.1typical
Continuity & Emotion ?80.1 ± 1.3± 1.5typical
YouTube Best Practices ?79.6 ± 2.5± 2.4typical
Hook Strength ?80.0 ± 2.9± 2.7typical
Length Adherence ?93.1 ± 4.3± 6.6typical
Slop Score ?83.3 ± 1.8± 1.6typical
Visual Cue Quality ?74.6 ± 4.4± 2.9more variable than most

Cost, Time and Tokens per Script

Logged on every run; none of these figures affects the Writing Score. Time is the average time to one accepted script, retries included. Cost is one script at the API list price.

Time per Script
1.1 minfaster than the board median of 1.5 min
Prompt tokens in
13.5kstyle guide + brief + research packet
Tokens Out
16.7kscript + any reasoning tokens
Cost per Script
$0.012measured from actual billed tokens; varies ± 0.005 between drafts
List Price Used
$0.15 / $0.6input / output, per million tokens

Scores by Judge

Four judges score the same 50 stored drafts blind against the same rubric: three LLM judges from three labs, plus Jev, a typed judge from TypeSafe that rates each metric on five described quality levels. No single judge's taste decides the result. The Writing Score of 80.3 combines their scores with the parts computed in code (Length Adherence and half of the Slop Score). The highest and lowest judge are 14.1 points apart on its Writing Score. That is close to the board median of 13.4. Each judge also has its own scale, so every card shows that judge's score against its average across the board. How the panel works.

Claude Opus 5.5 (medium)
Anthropic
76.5
+7.9 against its board average of 68.7
GLM 5.3 Flash
Z.ai
88.8
+5.9 against its board average of 82.9
GPT-6.1 Sol (medium)
OpenAI
74.7
+2.8 against its board average of 71.9
Jev 1.13
TypeSafe
81.3
+7.3 against its board average of 74.0

Evaluation Setup

How this exact configuration was run and judged: route, model ID, dates, sampling and prompt.

Frequently Asked Questions

How Good Is DeepSeek V4.1 Flash (high) at Writing?

DeepSeek V4.1 Flash ranks #17 of 87 models on ToneBench through its best configuration, DeepSeek V4.1 Flash (max). This configuration is #48 of 151 configurations, in the upper third of the board, with a Writing Elo of 1882 and a Writing Score of 80.3/100. The ranks come from full YouTube scripts written in the voice of Louis-François Bouchard and judged blind, with the channel's finished script as the reference. The benchmark measures long-form scripted writing, so the result does not carry over to every kind of text.

Is DeepSeek V4.1 Flash (high) the Best DeepSeek Model for Writing?

Not quite. DeepSeek V4.1 Flash is 2nd of 5 DeepSeek models. The top one is DeepSeek V4.1 Flash 1017, through DeepSeek V4.1 Flash 1017 (max), about 142 Elo ahead of this setting.

Is DeepSeek V4.1 Flash (high) Good Value for the Money?

It costs about $0.012 per script. MiMo-V2.6-Flash (thinking on) ranks higher with an overlapping Elo range and costs less.

What Are the Strengths and Weaknesses of DeepSeek V4.1 Flash (high)?

Its best-ranked metric is Length Adherence (93.1, 34th of 151 configurations, +8.4 against the board median), and its lowest-ranked is Substance, Accuracy & Value (80.2, 89th of 151, -1.7). The Scores by Metric section shows the board median and the board's best for each one.

How Was DeepSeek V4.1 Flash (high) Evaluated?

It was run through OpenRouter, with the exact model/route ID deepseek/deepseek-v4.1-flash at reasoning effort high, first run on 2026-10-08. It wrote each script 5 times with the provider's default sampling and no fine-tuning. The Evaluation Setup section and the methodology page have the details.

Does DeepSeek V4.1 Flash (high) Have Open Weights?

Yes. Its weights are public, so it can be downloaded and run on your own hardware. Check the license if you need open source in the strict sense. Among open-weight configurations it is 9th of 43 for writing.

← All models   ← Leaderboard