Thinking Levels · Anthropic

Claude Haiku 5.5: Writing Quality by Reasoning Effort

Last updated October 12, 2026, from board data generated at 01:13 UTC.

This page compares Claude Haiku 5.5 at 5 reasoning-effort settings, each tested on the same 10 scripts: the Writing Score (the 0 to 100 consensus of the judges and code), the Writing Elo (a rating from pairwise comparisons of those scores), and the cost and time of each setting. As a model, Claude Haiku 5.5 ranks #16 of 78 models through its best setting. The other ranks on this page are configuration ranks: one setting's place among all 127 configurations.

Highest Writing Elo

max#16 of 78 models · Elo 2009 · $0.103 per script

Lowest to Highest Effort

+505 Elolow to max · 23.0 times the cost

What More Effort Changes for Claude Haiku 5.5

More effort pays off well on Claude Haiku 5.5, and at Haiku prices the extra cost stays small. From low to max, it climbs 505 Elo, from #74 on the board to #31, enough for max to beat low on about 95% of scripts. The biggest step is high to xhigh, where the scripts start landing near their target length and the cues get better. You pay in time: max takes about 12.6 minutes a script, against 0.5 at low.

Recommended Setting

I'd run it at xhigh. That's where the big jump lands, at #41 for $0.029 a script and about 3.6 minutes. Max scores a little higher, but the two ranges still touch, and you'd pay 3.5 times the price and wait 12.6 minutes for a gain the board can't quite confirm.

The low settings are for bulk work where speed is all you care about. Just know what you give up: low and medium score clearly below xhigh, mostly on length, then on cues and the YouTube best-practices score. And if cheap is the goal, look at DeepSeek V4.1 Flash (max) too. It ranks above every Haiku setting, with a range clear of xhigh's, for $0.017 a script.

Every Setting on the Chart

Each point is one effort setting of Claude Haiku 5.5, joined from the lowest setting to the highest. Higher and further left is better: a higher score for less money or less time.

How to Use the Chart

A hollow marker is the provider default: the level the provider applies when the request sets no effort level.
The tabs choose the score on the vertical axis, X axis switches between cost and time, and Table lists the same numbers as rows.
Select a point for its numbers and a link to its scorecard.

Score on the vertical axis

View

X axis

On a small screen, Table lists the same numbers as rows.

All Effort Settings, Step by Step

Score colors:82 and up65 to 81.9under 65

Setting	Writing Elo	Step	Writing Score	Cost / Script	Time / Script	Configuration Rank
low	1504 ±44	-	75.2	$0.0045	0.5 min	#74
medium	1620 ±49	+116	76.7	$0.0067	0.8 min	#62
high	1723 ±40	+103	78.0	$0.012	1.5 min	#53
xhigh	1927 ±41	+204	80.0	$0.029	3.6 min	#41
max	2009 ±36	+82	80.8	$0.103	12.6 min	#31

Other Anthropic Models by Effort

Claude Opus 5.5

Best at xhigh, #1 of 78 models · +295 Elo, low to max

Claude Sonnet 5.5

Best at max, #2 of 78 models · +753 Elo, low to max

Claude Fable 5.1

Best at max, #3 of 78 models · +201 Elo, provider default to max

Claude Opus 5

Best at max, #5 of 78 models · +91 Elo, provider default to max

Claude Fable 5

Best at max, #6 of 78 models · +83 Elo, provider default to max

Claude Opus 4.8

Best at max, #15 of 78 models · +145 Elo, provider default to max

Claude Sonnet 5

Best at max, #23 of 78 models · +207 Elo, provider default to max

Writing Elo comes from comparing every pair of models on every script, which puts these settings on the same scale as the leaderboard. Step is the Writing Elo change from the previous explicit setting. Cost is one full script at API list prices, and time is the average for one accepted script, retries included. The thinking-levels page covers the other models, and the Anthropic family page lists every Anthropic model side by side.