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

All Models

373 models considered · scored on Intelligence Index

Last updated

Metric
Best Value Pick

DeepSeek V4 Flash 0731 (Reasoning, Max Effort)

49.9Intelligence Index
$0.17/1M tokens

The sharpest bend in the price-to-score curve out of 373 models — extra dollars beyond this point buy noticeably less extra capability.

Highlights

Intelligence

Artificial Analysis Intelligence Index · Higher is better

Claude Opus 5 (Max Effort)60.7Claude Fable 5 (Max Effort)59.9GPT-5.6 Sol (Max Effort)58.9Kimi K3 (Max Effort)57.1Claude Opus 4.8 (Max Effort)55.7GPT-5.6 Terra (Max Effort)55GPT-5.5 (Xhigh Effort)54.8Grok 4.5 (High Effort)53.8Claude Opus 4.7 (Max Effort)53.5Claude Sonnet 5 (Max Effort)53.4GLM-5.2 (Max Effort)51.1Muse Spark 1.1 (Xhigh Effort)50.6Gemini 3.5 Flash (High Effort)50.2
Opus 5 (Max)
Fable 5 (Max)
5.6 Sol (Max)
K3 (Max)
Opus 4.8 (Max)
5.6 Terra (Max)
5.5 (XH)
4.5 (H)
Opus 4.7 (Max)
Sonnet 5 (Max)
5.2 (Max)
Muse Spark 1.1 (XH)
3.5 Flash (H)

Bars past the dashed line are top models from competitive makers that didn't crack the strict top 10 on this metric - shown so near-flagship providers (and your subscription makers) stay visible for comparison.

Price

Blended $ per 1M tokens (3:1 in:out) · Lower is better

Grok 4.5 (High Effort)$3Claude Sonnet 5 (Max Effort)$4GPT-5.6 Terra (Max Effort)$4.50Kimi K3 (Max Effort)$6Claude Opus 5 (Max Effort)$10Claude Opus 4.8 (Max Effort)$10Claude Opus 4.7 (Max Effort)$10GPT-5.6 Sol (Max Effort)$11.25GPT-5.5 (Xhigh Effort)$11.25Claude Fable 5 (Max Effort)$20
4.5 (H)
Sonnet 5 (Max)
5.6 Terra (Max)
K3 (Max)
Opus 5 (Max)
Opus 4.8 (Max)
Opus 4.7 (Max)
5.6 Sol (Max)
5.5 (XH)
Fable 5 (Max)

Speed

Output tokens per second · Higher is better

GPT-5.6 Terra (Max Effort)128Claude Sonnet 5 (Max Effort)79GPT-5.6 Sol (Max Effort)67Claude Fable 5 (Max Effort)60Claude Opus 5 (Max Effort)56Grok 4.5 (High Effort)52Kimi K3 (Max Effort)34
5.6 Terra (Max)
Sonnet 5 (Max)
5.6 Sol (Max)
Fable 5 (Max)
Opus 5 (Max)
4.5 (H)
K3 (Max)
Sort by
Show
Budget
≥50%Balanced
≥80%Premium
≥90%
≥ 80% of top score
#ModelValueIntelligencePrice
1Grok 4.5 (High Effort)
100		53.8$3
2Hy3-preview (Reasoning) Frontier
100		33.6$0.11
3DeepSeek V4 Flash 0731 (Max Effort) Best Value
100		49.9$0.17
4GPT-5.6 Luna (Max Effort) Frontier
100		51.2$0.45
5Muse Spark 1.1 (Xhigh Effort) Featured
100		50.6$2
6GLM-5.2 (Max Effort) Featured
100		51.1$2.15
7Gemini 3.5 Flash (High Effort) Featured
83		50.2$3.38
8Claude Sonnet 5 (Max Effort)
74		53.4$4
9GPT-5.6 Terra (Max Effort)
68		55$4.50
10Kimi K3 (Max Effort)
53		57.1$6
11Claude Opus 5 (Max Effort)
34		60.7$10
12Claude Opus 4.8 (Max Effort)
31		55.7$10
13Claude Opus 4.7 (Max Effort)
30		53.5$10
14GPT-5.6 Sol (Max Effort)
29		58.9$11.25
15GPT-5.5 (Xhigh Effort)
27		54.8$11.25
16Claude Fable 5 (Max Effort)
17		59.9$20
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 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 Show: Budget / Balanced / Premium to control how far down that list goes - Balanced (default) only shows models at least 80% as capable as the top model; Budget loosens that to 50%; Premium tightens it to 90%. 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.