Model Comparison
Claude Sonnet 5.5 vs GPT-5.6 Luna
Claude Sonnet 5.5 (max effort) leads GPT-5.6 Luna (ultra) by 749 Elo, and their 95% ranges sit 646 Elo apart. Among 78 models, each ranked by its best configuration, Claude Sonnet 5.5 is #2 and GPT-5.6 Luna is #24. Cost per script: $1.38 for Claude Sonnet 5.5 (max effort) and $0.026 for GPT-5.6 Luna (ultra).
Last updated , from board data generated at 01:13 UTC.
Picks and Verdict
PickClaude Sonnet 5.5
Claude Sonnet 5.5 (xhigh)
Pick Claude Sonnet 5.5 (xhigh) for scripts you'll publish, since it sits 536 Elo above GPT-5.6 Luna (ultra) for $0.302 a script.
PickGPT-5.6 Luna
GPT-5.6 Luna (ultra)
Pick GPT-5.6 Luna (ultra) for cheap first drafts that hold their length and carry solid substance, at $0.026 a script.
Claude Sonnet 5.5, the #2 model of 78, is far ahead of GPT-5.6 Luna, the #24 model, at the top: its max effort setting leads GPT-5.6 Luna (ultra) by 749 Elo, with 95% ranges far apart. The cheap end is much closer. Luna at ultra lands between Claude Sonnet 5.5 (medium) and Claude Sonnet 5.5 (low), and its range overlaps both, so the board can't separate it from Sonnet's two cheapest settings. Luna charges $0.026 a script for that, less than half of Sonnet at medium, though it takes over three minutes where Sonnet needs under one. It needs ultra to get there. GPT-5.6 Luna (medium) sits 203 Elo lower with a separate range, so I wouldn't run it. Luna does score higher on substance and holds the target length far better than Sonnet at medium. Voice is the catch: it trails by 5.3 points on Tone & Voice Match, the metric I weigh most, and I'd pay the extra four cents before I'd rewrite for tone. For a script I'll record, I'd go up to Claude Sonnet 5.5 (xhigh), 536 Elo above Luna's best for $0.302.
Disclosure: Claude Sonnet 5.5 and GPT-5.6 Luna each come from the same lab as one of the judges (Claude Opus 5.5 and GPT-6.1 Sol). Neither is 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 Sonnet 5.5 and GPT-5.6 Luna, scored on the same 10 scripts by the same four judges.
- 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.
- When two bars overlap, the board cannot order those two configurations. Bars far apart are strong evidence for the order.
- 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).
Shaded rows are the configurations chosen for the metric comparison below.
| Configuration | Writing Elo | 95% Range | Writing Score | Cost / Script | Time / Script | Configuration Rank |
|---|---|---|---|---|---|---|
| Claude Sonnet 5.5 (max effort) | 2561 | 2504 to 2614 | 84.6 | $1.38 | 14.3 min | #3 |
| Claude Sonnet 5.5 (xhigh) | 2348 | 2304 to 2383 | 83.2 | $0.302 | 3.1 min | #10 |
| Claude Sonnet 5.5 (high) | 2040 | 1984 to 2087 | 81.0 | $0.092 | 1.0 min | #27 |
| Claude Sonnet 5.5 (medium) | 1856 | 1812 to 1900 | 79.4 | $0.065 | 0.6 min | #47 |
| GPT-5.6 Luna (ultra) | 1812 | 1768 to 1858 | 78.8 | $0.026 | 3.1 min | #49 |
| Claude Sonnet 5.5 (low) | 1808 | 1771 to 1843 | 78.9 | $0.061 | 0.7 min | #50 |
| GPT-5.6 Luna (medium) | 1609 | 1572 to 1647 | 76.7 | $0.0065 | 0.5 min | #66 |
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.
Computed Findings Across Thinking Levels
Each finding is computed from the configurations above. Where two 95% ranges overlap, the board cannot order those two configurations.
- Best Configurations Head to Head. Claude Sonnet 5.5 (max effort) leads GPT-5.6 Luna (ultra) by 749 Elo, and their 95% ranges do not overlap (2504 to 2614 against 1768 to 1858), which is strong evidence for that order.
- Lowest Thinking Levels Head to Head. At their lowest thinking levels, Claude Sonnet 5.5 (low) at Elo 1808 and $0.061 per script leads GPT-5.6 Luna (medium) at Elo 1609 and $0.0065 per script by 199 Elo. Their 95% ranges do not overlap.
- Same Level Name, Side by Side. At medium, the one level name both models use, Claude Sonnet 5.5 has the higher Writing Elo: 1856 against 1609. Their 95% ranges do not overlap there. Level names come from each provider, so the same name does not promise the same reasoning budget.
- Cheapest Claude Sonnet 5.5 Configuration Above GPT-5.6 Luna's Best. The cheapest Claude Sonnet 5.5 configuration with a higher Writing Elo than GPT-5.6 Luna (ultra), at Elo 1812 and $0.026 per script, is Claude Sonnet 5.5 (medium), at Elo 1856 and $0.065 per script. Their 95% ranges overlap, so the board cannot separate them. It costs about 2.5 times as much per script.
- Lowest to Highest Thinking Level. Claude Sonnet 5.5 moves +753 Elo from low to max effort, at about 23 times the cost per script. GPT-5.6 Luna moves +203 Elo from medium to ultra, at about 4 times the cost per script.
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 Sonnet 5.5 (max effort) scores higher on all nine metrics. Weighted by each metric's share of the Writing Score, Tone & Voice Match does the most to put Claude Sonnet 5.5 (max effort) ahead, with a 8.6-point lead on that metric. The weight beside each metric is its share of the Writing Score.
Consistency Across Drafts
| Claude Sonnet 5.5 (max effort) | GPT-5.6 Luna (ultra) | |
|---|---|---|
| Score spread across drafts (±) | 1.22 | 2.07 |
| Lowest script score | 83.0 Script 10 | 76.0 Script 7 |
| Highest script score | 86.7 Script 9 | 80.8 Script 9 |
| Range across scripts | 3.7 | 4.8 |
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 Sonnet 5.5 (max effort) minus GPT-5.6 Luna (ultra), and the higher score is in bold.
| Script | Claude Sonnet 5.5 (max effort) | GPT-5.6 Luna (ultra) | Gap |
|---|---|---|---|
| Script 1 opinion / warning explainer | 84.2 | 78.5 | +5.7 |
| Script 2 news-analysis / skeptical explainer | 84.8 | 80.3 | +4.5 |
| Script 3 personal roadmap / opinion | 84.0 | 79.7 | +4.3 |
| Script 4 short explainer | 85.0 | 77.4 | +7.6 |
| Script 5 news-analysis / opinion explainer | 84.8 | 79.7 | +5.1 |
| Script 6 founder announcement / personal origin story | 84.6 | 80.6 | +4.0 |
| Script 7 personal technical walkthrough / agentic workflow case study | 85.0 | 76.0 | +9.0 |
| Script 8 product release announcement / personal observation | 83.9 | 78.4 | +5.5 |
| Script 9 career guide / hiring analysis | 86.7 | 80.8 | +5.9 |
| Script 10 engineering process walkthrough / presentation adaptation | 83.0 | 76.9 | +6.1 |
Scorecards and Thinking Pages
Each configuration name in the table above opens its own scorecard.
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