All Models

Best-value picks among All Models models, ranked live from artificialanalysis.ai.

82 models ranked · scored on Intelligence Index

Last updated 2026-10-30 05:46 UTC

Metric

Capability

Every model, nothing filtered out

Best Value Pick

Claude Haiku 5.5 (Max)

36.2Intelligence Index

$0.142per task (est.)

$0.20/1M tokens

The sharpest bend in the price-to-score curve out of 82 models — extra dollars beyond this point buy noticeably less extra capability.

Highlights

Intelligence

Artificial Analysis Intelligence Index · Higher is better

4 Carbon (H)

Opus 5.5 (Max)

Opus 5.5 (XH)

6 Astra (Max)

6.1 Sol (Max)

Sonnet 5.5 (Max)

Opus 5.5 (H)

6.1 Terra (Max)

4 Argon (H)

Fable 5.1 (XH)

Value

score ÷ cost per task · Higher is better

Haiku 5.5 (Max)

6.1 Terra (Max)

4 Carbon (H)

6.1 Sol (Max)

Sonnet 5.5 (Max)

Opus 5.5 (H)

4 Argon (H)

Opus 5.5 (XH)

6 Astra (Max)

Opus 5.5 (Max)

Fable 5.1 (XH)

Cost per task

price × tokens used · Lower is better

Haiku 5.5 (Max)

6.1 Terra (Max)

4 Carbon (H)

Opus 5.5 (H)

4 Argon (H)

6.1 Sol (Max)

Sonnet 5.5 (Max)

Opus 5.5 (XH)

6 Astra (Max)

Opus 5.5 (Max)

Fable 5.1 (XH)

Tokens per task

cost per task ÷ price · Lower is better

6 Astra (Max)

4 Carbon (H)

Opus 5.5 (H)

4 Argon (H)

Opus 5.5 (XH)

6.1 Sol (Max)

Sonnet 5.5 (Max)

6.1 Terra (Max)

Fable 5.1 (XH)

Haiku 5.5 (Max)

Opus 5.5 (Max)

Price

(3 × input + output) ÷ 4 · The number that misleads

Haiku 5.5 (Max)

6.1 Terra (Max)

6.1 Sol (Max)

Sonnet 5.5 (Max)

4 Carbon (H)

Opus 5.5 (Max)

Opus 5.5 (XH)

Opus 5.5 (H)

4 Argon (H)

6 Astra (Max)

Fable 5.1 (XH)

Speed

Output tokens per second · Higher is better

Haiku 5.5 (Max)

Sonnet 5.5 (Max)

Opus 5.5 (Max)

6.1 Terra (Max)

6 Astra (Max)

4 Carbon (H)

Opus 5.5 (XH)

6.1 Sol (Max)

Opus 5.5 (H)

4 Argon (H)

Fable 5.1 (XH)

#	Model	Value score ÷ cost/task	Intelligence	Cost / task price × tokens	Price $ / 1M
1	Claude Haiku 5.5 (Max Effort) Best Value	100	36.2	$0.142 ~	$0.20
2	GPT-6.1 Terra (Max Effort)	54	51	$1.42 ~	$2
3	Gemini 4 Carbon (High Effort)	42	59.4	$2.75 ~	$8
4	GPT-6.1 Sol (Max Effort)	40	54.5	$2.83 ~	$4
5	Claude Sonnet 5.5 (Max, Default Fallback) (Max)	40	54.2	$2.83 ~	$4
6	Claude Opus 5.5 (High, Default Fallback) (High)	39	51.7	$2.75 ~	$8
7	Gemini 4 Argon (High Effort)	39	50.4	$2.75 ~	$8
8	Claude Opus 5.5 (Xhigh, Default Fallback) (Xhigh)	25	57	$5.67 ~	$8
9	GPT-6 Astra (Max Effort)	22	56.9	$6.24	$20
10	Claude Opus 5.5 (Max, Default Fallback) (Max)	12	58.4	$10.04	$8
11	Claude Fable 5.1 (Xhigh, Default Fallback) (Xhigh)	0	49.7	$14.17 ~	$20
No models made this category's table - every one of its models was beaten on both price and intelligence by another maker's model here, so none reached the top 10 or the value frontier. No per-company quota forces a spot - try a different category.

How this page works

Highlights shows the true top 10 models by intelligence - no per-company limit, so if one company genuinely has several of the top models right now, several of its models are shown. Ranked three ways - Intelligence, Price, and Speed - so you can see how the same lineup trades off against each other. Each model keeps one color across all three charts. Value (0-100) below that shows how much intelligence you get per dollar, scaled so the best value model shown scores 100 - tables sort by Value by default; click "Intelligence" or "Name" above one to sort by those instead. The three calculated columns are, in full: Cost / task = price per 1M tokens × tokens the model actually uses ÷ 1,000,000 Value = intelligence score ÷ cost per task Price = (3 × input + output) ÷ 4, per 1M tokens The Best Value Pick is the sharpest bend in the price-to-intelligence curve across every model in the category. The table also includes every other model on that same efficiency frontier (tagged Frontier) even if it didn't score high enough to crack the top 10 - a cheap, well-balanced model would otherwise be invisible no matter how the table is sorted. Use the Capability buttons to control how far down that list goes - All (default) filters nothing; Flagship keeps models within 10% of the best score, Strong within 20%, Broad within 50%. The Best Value Pick always stays visible regardless of which tier is selected. Click Update on a category page (or Update All on the home page) to refetch live data any time.