Towards AITowards AIToneBench

Model Comparison

Claude Opus 5 vs DeepSeek V4.1 Flash

Claude Opus 5 (max effort) leads DeepSeek V4.1 Flash (max) by 299 Elo, and their 95% ranges sit 202 Elo apart. Even Claude Opus 5's lowest-rated configuration, Claude Opus 5 (high), ranks above DeepSeek V4.1 Flash (max). Among 87 models, each ranked by its best configuration, Claude Opus 5 is #6 and DeepSeek V4.1 Flash is #17. Cost per script: $0.736 for Claude Opus 5 (max effort) and $0.017 for DeepSeek V4.1 Flash (max).

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

Claude Opus 5
2255 Writing Elo of Claude Opus 5 (max effort), its best configuration#6 of 87 models · 83.3/100 Writing Score · $0.736 per script2 configurations on the board
DeepSeek V4.1 Flash
1956 Writing Elo of DeepSeek V4.1 Flash (max), its best configuration#17 of 87 models · 81.0/100 Writing Score · $0.017 per script4 configurations on the board

Picks and Verdict

PickClaude Opus 5

Claude Opus 5 (max effort)

Pick Claude Opus 5 (max effort) when the script matters, since it leads DeepSeek's best by 299 Elo at $0.736 a script.

PickDeepSeek V4.1 Flash

DeepSeek V4.1 Flash (high)

Pick DeepSeek V4.1 Flash (high) for bulk drafting at $0.012 a script, a setting whose range overlaps max's.

Every Claude Opus 5 setting rates above every DeepSeek V4.1 Flash setting, so this page comes down to price. Claude Opus 5 is the #6 model of 87 and DeepSeek V4.1 Flash the #17 model, each ranked by its best setting. Claude Opus 5 (max effort) leads DeepSeek V4.1 Flash (max) by 299 Elo with separate 95% ranges, and it costs $0.736 a script against $0.017, about 43 times as much. Neither model's top step has separate ranges. On Opus, max effort adds 108 Elo over high for about 1.6 times the cost, and the two ranges overlap at the edges. On DeepSeek, thinking is worth 394 Elo across its ladder, from no thinking to max, but the top two ranges overlap, and DeepSeek V4.1 Flash (high) costs $0.012 and finishes in about a minute. Both of those DeepSeek settings land closer to the requested length than either Opus setting, and DeepSeek at max gives steadier drafts than Opus at high. I'd still pay for Opus at max effort when the script matters, and run DeepSeek at high when I need lots of drafts for almost nothing.

Disclosure: Claude Opus 5 comes from the same lab as one of the four judges, Claude Opus 5.5. The other three sit outside its lab, so it is never scored by its own lab alone (how the panel works). The judge analysis measures how much each judge favors its own lab.

How to Read the Writing Score, Writing Elo and 95% Ranges

Writing Score and Writing Elo. The Writing Score (0 to 100) is the weighted average of the nine metric scores: seven from the judges, Length Adherence from code and the Slop Score half from each. The Writing Elo is a rating from pairwise comparisons: on each script, each configuration's average across its drafts is compared with every other configuration's, and gaps within the draft-to-draft variation count as draws. Elo rewards winning often, so the two can put configurations in a different order.

Elo ranges. Each 95% range shows how far a configuration's Elo could move if the drafts had come out slightly differently. It comes from resampling every configuration's recorded draft scores on each reference script and ranking again. Comparing two separate ranges does not test the Elo gap between them: ranges far apart are strong evidence for the order, and overlapping ranges do not mean a tie. The reference scripts stay fixed, so the ranges say nothing about new scripts. The methodology page has the full description.

All Thinking Levels of Both Models

Every ranked configuration of Claude Opus 5 and DeepSeek V4.1 Flash, scored on the same 10 scripts by the same four judges.

How to Read the Chart
  • Each row is one configuration: the model at one thinking level. The dot is its Writing Elo, a rating from pairwise comparisons of the scores on each script, and the bar is its 95% range: how far that Elo could move if the drafts had come out slightly differently.
  • Each bar describes one configuration. Overlap alone does not establish a tie or decide their order. Comparing these ranges does not directly test the difference between a pair.
  • The table under the chart lists the same configurations in the same order and adds the Writing Score (the 0 to 100 consensus of the judges and code), cost, time and configuration rank (the place among all configurations on the board).
Claude Opus 5DeepSeek V4.1 Flash
ConfigurationWriting Elo95% RangeWriting ScoreCost / ScriptTime / ScriptReasoning TokensConfiguration Rank
Claude Opus 5 (max effort)22552197 to 230583.3$0.7365.5 minN/A#16
Claude Opus 5 (high)21472097 to 220182.6$0.4723.5 minN/A#20
DeepSeek V4.1 Flash (max)19561914 to 199581.0$0.0172.4 min22,010#36
DeepSeek V4.1 Flash (high)18821842 to 192180.3$0.0121.1 min13,809#48
DeepSeek V4.1 Flash (low)17841725 to 183979.3$0.00770.7 min6,740#60
DeepSeek V4.1 Flash (no thinking)15621514 to 160576.8$0.00370.3 min0#82

Cost is one script at the API list price of the route used, uncached and without judging. Time is the average for one accepted script, retries included. N/A under Reasoning Tokens means the board has no separate reasoning count for that route.

Computed Findings Across Thinking Levels

Each finding is computed from the configurations above. The 95% ranges describe individual configurations; overlap alone establishes neither a tie nor an order for a pair.

Writing Elo at Each Shared Level Name

LevelClaude Opus 5DeepSeek V4.1 FlashGap95% Ranges
high21471882+265Separate
max22551956+299Separate

Gap is Claude Opus 5 minus DeepSeek V4.1 Flash at the same level name, using each model's highest-rated configuration at that level.

Two Chosen Configurations Compared

Pick one configuration of each model to compare their metric scores, consistency across drafts and per-script scores. Each menu starts on the model's best configuration.

Scores by Metric

Claude Opus 5 (max effort)DeepSeek V4.1 Flash (max)Solid bar: higher score in that row

Claude Opus 5 (max effort) scores higher on eight of the nine metrics and DeepSeek V4.1 Flash (max) on one. Weighted by each metric's share of the Writing Score, Tone & Voice Match does the most to put Claude Opus 5 (max effort) ahead, with a 2.7-point lead on that metric. The weight beside each metric is its share of the Writing Score.

Tone & Voice Match19% weight
80.8
78.1
Writing Craft & Clarity13% weight
80.5
78.0
Substance, Accuracy & Value15% weight
84.5
82.5
Continuity & Emotion14% weight
83.7
80.1
YouTube Best Practices12% weight
83.5
79.8
Hook Strength10% weight
82.3
80.1
Length Adherence8% weight
92.0
95.9
Slop Score5% weight
85.1
83.2
Visual Cue Quality4% weight
79.8
76.0

Consistency Across Drafts

Claude Opus 5 (max effort)DeepSeek V4.1 Flash (max)
Score spread across drafts (±)1.101.30
Lowest script score81.9 Script 1079.7 Script 10
Highest script score84.5 Script 782.4 Script 7
Range across scripts2.62.7

Spread is how much the Writing Score moves between a configuration's drafts, and lower is steadier. Green marks the better number in each row.

Scores by Script

Writing Score on each reference script. Gap is Claude Opus 5 (max effort) minus DeepSeek V4.1 Flash (max), and the higher score is in bold.

ScriptClaude Opus 5 (max effort)DeepSeek V4.1 Flash (max)Gap
Script 1 opinion / warning explainer83.080.8+2.2
Script 2 news-analysis / skeptical explainer83.481.2+2.2
Script 3 personal roadmap / opinion83.281.2+2.0
Script 4 short explainer83.280.1+3.1
Script 5 news-analysis / opinion explainer83.381.1+2.2
Script 6 founder announcement / personal origin story83.480.8+2.6
Script 7 personal technical walkthrough / agentic workflow case study84.582.4+2.1
Script 8 product release announcement / personal observation83.080.7+2.3
Script 9 career guide / hiring analysis83.682.2+1.4
Script 10 engineering process walkthrough / presentation adaptation81.979.7+2.2

Compare With Another Model

Pick a second model. A pair with its own written match-up opens that page, and any other pair opens in the comparison tool.

Nearby Match-Ups by Best Writing Elo

Every configuration wrote five drafts for each of the same 10 briefs, from the same prompt. Generation was closed-book, with the prompt as the only source and no web search or other tools. Each judge scores every draft once: Claude Opus 5.5 at medium effort through the Claude Code subscription, GPT-6.1 Sol at medium reasoning through the Codex CLI, GLM 5.3 Flash through OpenRouter, and Jev 1.13, a typed judge from TypeSafe that answers each metric as a multiple-choice question. The channel's finished script is their reference for the voice, and it gets no score on the board itself.

← All comparisons