AI Tools Oasis Logo
AI Tools Oasis

Discover Smart Tools

Live Data from Artificial Analysis
Last updated: Jul 26, 2026, 7:06 AM

AI Intelligence IndexAI Models Benchmark

The most comprehensive platform for comparing and analyzing AI models.
Intelligence • Speed • Cost

586+AI Models
HourlyUpdates
CompleteComparison
56+Global Providers
Avg. Intelligence
18.78
points
Lowest Price
$0.0200
per million
Fastest Model
Mercury 2
1119 tok/s
Lowest Latency
Command A+
0.17s
Smart Advisor

AI Model Advisor

Let our smart tool help you pick the best model for your needs by analyzing your requirements against our database.

What is your main priority?

1 of 4

Choose what matters most to you in a model

Model Comparison Summary

Intelligence

Claude Opus 5 (Adaptive Reasoning, Max Effort)

The smartest model at 60.7

Output Speed

Mercury 2

Fastest at 1119 tok/s

Latency

Command A+

Lowest latency at 0.17s

Price

Gemma 4 E4B (Non-reasoning)

Cheapest at $0.02 per million tokens

Context Window

—

Data not available

Highlights

Intelligence (Quality Index)

Aggregate quality index; higher is better

Speed

Output tokens per second; higher is better

Price (Input)

USD per million tokens; lower is better

Detailed Model View

Explore each model's details with comprehensive benchmark results and interactive charts.

C

Claude Opus 5 (Adaptive Reasoning, Max Effort)

Anthropic

Intelligence: 60.7Speed: 44 t/sPrice: $5.00

C

Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)

Anthropic

Intelligence: 60.1Speed: 60 t/sPrice: $5.00

C

Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)

Anthropic

Intelligence: 59.9Speed: 58 t/sPrice: $10.00

G

GPT-5.6 Sol (max)

OpenAI

Intelligence: 58.9Speed: 74 t/sPrice: $5.00

C

Claude Opus 5 (Adaptive Reasoning, High Effort)

Anthropic

Intelligence: 58.9Speed: 50 t/sPrice: $5.00

G

GPT-5.6 Sol (xhigh)

OpenAI

Intelligence: 57.7Speed: 64 t/sPrice: $5.00

K

Kimi K3

Kimi

Intelligence: 57.1Speed: 33 t/sPrice: $3.00

C

Claude Opus 5 (Adaptive Reasoning, Medium Effort)

Anthropic

Intelligence: 56.3Speed: 56 t/sPrice: $5.00

G

GPT-5.6 Sol (high)

OpenAI

Intelligence: 55.9Speed: 66 t/sPrice: $5.00

C

Claude Opus 4.8 (Adaptive Reasoning, Max Effort)

Anthropic

Intelligence: 55.7Speed: 63 t/sPrice: $5.00

G

GPT-5.6 Terra (max)

OpenAI

Intelligence: 55.0Speed: 128 t/sPrice: $2.50

G

GPT-5.5 (xhigh)

OpenAI

Intelligence: 54.8Speed: 0 t/sPrice: $5.00

G

Grok 4.5 (high)

SpaceXAI

Intelligence: 53.8Speed: 56 t/sPrice: $2.00

G

GPT-5.6 Sol (medium)

OpenAI

Intelligence: 53.6Speed: 60 t/sPrice: $5.00

C

Claude Opus 4.7 (Adaptive Reasoning, Max Effort)

Anthropic

Intelligence: 53.5Speed: 0 t/sPrice: $5.00

3D Dimensions

Beyond two dimensions: a three-variable analysis combining price, performance, and speed in one view.

Price vs. Intelligence

Size = Speed (Tokens/Sec). Find the ideal balance.

(Zoom: 100%)
Use the zoom controls to adjust the chart view
How to read: Bubble size representsSpeed. The larger the bubble, the higher the value.

Latency vs. Speed

Size = Intelligence. Technical vs. cognitive performance.

(Zoom: 100%)
Use the zoom controls to adjust the chart view
How to read: Bubble size representsIntelligence. The larger the bubble, the higher the value.

Correlation Analysis

Explore the hidden relationships between price, speed, and intelligence to surface valuable insights about model efficiency.

Intelligence vs. Price

Does higher quality always require paying more?

(Zoom: 100%)
Use the zoom controls to adjust the chart view

Comprehensive Heatmap

A quick overview revealing each model's strengths and weaknesses. Green always means best (even for price and latency).

Matrix View(Zoom: 100%)
Model
Intelligence
Score
Speed
tok/s
Latency
ms
Price
$/M
Context
k tok
Claude Opus 5 (Adaptive Reasoning, Max Effort)
Anthropic
60.7
43.9
28.698s
$5.00
0k
Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)
Anthropic
60.1
60.4
22.558s
$5.00
0k
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
Anthropic
59.9
58.3
52.464s
$10.00
0k
GPT-5.6 Sol (max)
OpenAI
58.9
73.9
86.493s
$5.00
0k
Claude Opus 5 (Adaptive Reasoning, High Effort)
Anthropic
58.9
50.1
10.052s
$5.00
0k
GPT-5.6 Sol (xhigh)
OpenAI
57.7
64.2
31.000s
$5.00
0k
Kimi K3
Kimi
57.1
33.1
102.958s
$3.00
0k
Claude Opus 5 (Adaptive Reasoning, Medium Effort)
Anthropic
56.3
55.6
9.657s
$5.00
0k
GPT-5.6 Sol (high)
OpenAI
55.9
65.8
11.263s
$5.00
0k
Claude Opus 4.8 (Adaptive Reasoning, Max Effort)
Anthropic
55.7
62.5
46.471s
$5.00
0k
GPT-5.6 Terra (max)
OpenAI
55.0
128.0
110.138s
$2.50
0k
GPT-5.5 (xhigh)
OpenAI
54.8
0.0
0.110s
$5.00
0k
Grok 4.5 (high)
SpaceXAI
53.8
55.8
12.627s
$2.00
0k
GPT-5.6 Sol (medium)
OpenAI
53.6
60.0
5.280s
$5.00
0k
Claude Opus 4.7 (Adaptive Reasoning, Max Effort)
Anthropic
53.5
0.0
0.290s
$5.00
0k
Claude Sonnet 5 (Adaptive Reasoning, Max Effort)
Anthropic
53.4
83.4
108.375s
$2.00
0k
GPT-5.5 (high)
OpenAI
53.1
0.0
0.110s
$5.00
0k
GPT-5.6 Terra (xhigh)
OpenAI
51.6
120.0
7.701s
$2.50
0k
GPT-5.4 (xhigh)
OpenAI
51.4
0.0
0.100s
$2.50
0k
GPT-5.6 Luna (max)
OpenAI
51.2
171.4
86.827s
$1.00
0k
GLM-5.2 (max)
Z AI
51.1
156.7
0.906s
$1.40
0k
Muse Spark 1.1 (xhigh)
Meta
50.6
123.9
1.023s
$1.25
0k
Claude Opus 5 (Adaptive Reasoning, Low Effort)
Anthropic
50.6
46.7
2.763s
$5.00
0k
GPT-5.5 (medium)
OpenAI
50.4
0.0
0.360s
$5.00
0k
Gemini 3.5 Flash (high)
Google
50.2
250.3
17.160s
$1.50
0k
Gemini 3.6 Flash (high)
Google
50.1
219.4
14.515s
$1.50
0k
GPT-5.6 Sol (low)
OpenAI
49.4
69.3
2.798s
$5.00
0k
GPT-5.6 Luna (xhigh)
OpenAI
49.1
161.2
36.093s
$1.00
0k
GPT-5.6 Terra (high)
OpenAI
49.0
114.6
2.437s
$2.50
0k
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)
Anthropic
47.2
0.0
0.360s
$3.00
0k
Gemini 3.1 Pro Preview
Google
46.5
132.2
30.979s
$2.00
0k
GPT-5.6 Luna (high)
OpenAI
46.1
174.3
6.215s
$1.00
0k
Qwen3.7 Max
Alibaba
46.0
199.6
1.603s
$2.50
0k
GPT-5.6 Terra (medium)
OpenAI
45.6
116.0
1.781s
$2.50
0k
Gemini 3.5 Flash (medium)
Google
45.4
265.2
11.897s
$1.50
0k
MiniMax-M3
MiniMax
44.4
86.6
1.186s
$0.30
0k
GPT-5.3 Codex (xhigh)
OpenAI
44.3
125.8
53.330s
$1.75
0k
DeepSeek V4 Pro (Reasoning, Max Effort)
DeepSeek
44.3
70.9
1.021s
$0.43
0k
Kimi K2.6
Kimi
44.2
0.0
0.210s
$0.95
0k
Motif 3 (Beta)
Motif Technologies
44.1
0.0
0.170s
$0.00
0k
Claude Opus 4.6 (Adaptive Reasoning, Max Effort)
Anthropic
43.7
0.0
0.280s
$5.00
0k
GPT-5.5 (low)
OpenAI
43.5
0.0
0.130s
$5.00
0k
Muse Spark
Meta
43.1
0.0
0.450s
$0.00
0k
DeepSeek V4 Pro (Reasoning, High Effort)
DeepSeek
43.1
72.9
0.922s
$0.43
0k
Claude Opus 4.7 (Non-reasoning, High Effort)
Anthropic
42.7
46.3
1.520s
$5.00
0k
MiMo-V2.5-Pro
Xiaomi
42.2
65.2
2.111s
$0.43
0k
GPT-5.2 (xhigh)
OpenAI
42.2
0.0
0.480s
$1.75
0k
Kimi K2.7 Code
Kimi
41.9
44.6
1.277s
$0.95
0k
Claude Sonnet 5 (Non-reasoning, High Effort)
Anthropic
41.7
64.3
1.219s
$2.00
0k
GPT-5.6 Sol (Non-reasoning)
OpenAI
41.2
71.0
1.074s
$5.00
0k
Use the zoom controls to adjust the chart view
Poor
Fair
Good
Excellent

Quick Navigation

Intelligence Capabilities

Models ranked by cognitive and coding proficiency benchmarks.

Models ranked by overall Quality Index

Average model performance across a wide range of standard benchmarks (MMLU, GPQA, etc.)

30 models

Open-Source Models
Open Weights

Models whose weights are available to download and use freely

Closed Models (API)
Proprietary

Models available exclusively via APIs

Performance Overview

A comprehensive analysis linking intelligence, speed, and cost in one view.

Intelligence vs. Speed vs. Price

Bubble size represents value for money (the larger it is, the "cheaper" the model is relative to its performance).

X-axis: intelligence level
Y-axis: generation speed
C

Claude Opus 5 (Adaptive Reasoning, Max Effort)

Reasoning

Intel.

61

Speed

44

Price

$5

C

Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)

Reasoning

Intel.

60

Speed

60

Price

$5

C

Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)

Reasoning

Intel.

60

Speed

58

Price

$10

G

GPT-5.6 Sol (max)

Intel.

59

Speed

74

Price

$5

C

Claude Opus 5 (Adaptive Reasoning, High Effort)

Reasoning

Intel.

59

Speed

50

Price

$5

G

GPT-5.6 Sol (xhigh)

Intel.

58

Speed

64

Price

$5

K

Kimi K3

Intel.

57

Speed

33

Price

$3

C

Claude Opus 5 (Adaptive Reasoning, Medium Effort)

Reasoning

Intel.

56

Speed

56

Price

$5

Pricing & Value

A comparative analysis of model prices versus the value they deliver.

Input Price Comparison

USD per million tokens (Input) — lower is better

Value-for-Money Table

Value analysis (intelligence divided by price). Arrows indicate a "good deal".

ModelPrice ($)IntelligenceValue
Motif 3 (Beta)
$044
∞
Muse Spark
$043
∞
MiMo-V2-Pro
$040
∞
JT-4.1 Flash 236B A21B
$039
∞
GLM-5-Turbo
$038
∞
MiMo-V2-Omni-0327
$036
∞
MiMo-V2-Omni
$035
∞
GLM 5V Turbo (Reasoning)
$035
∞
LongCat 2.0
$034
∞
MiMo-V2-Flash (Feb 2026)
$033
∞
Qwen3 Max Thinking
$032
∞
Grok 4.1 Fast (Reasoning)
$031
∞
Gemma 4 31B (Reasoning)
$029
∞
JT-35B-Flash
$028
∞
KAT-Coder-Pro V1
$028
∞
Claude 3.7 Sonnet (Reasoning)
$027
∞
Doubao Seed Code
$026
∞
MiMo-V2-Flash (Non-reasoning)
$025
∞
Gemini 2.5 Flash Preview (Sep '25) (Reasoning)
$024
∞
Gemini 2.5 Pro Preview (Mar' 25)
$023
∞
Command A+
$023
∞
DeepSeek V3.2 Speciale
$022
∞
K-EXAONE (Reasoning)
$022
∞
ERNIE 5.0 Thinking Preview
$022
∞
Grok Code Fast 1
$022
∞
Apriel-v1.5-15B-Thinker
$021
∞
Apriel-v1.6-15B-Thinker
$021
∞
Qwen3.5 9B (Non-reasoning)
$020
∞
EXAONE 4.5 33B
$020
∞
North Mini Code
$020
∞
GLM-4.5 (Reasoning)
$020
∞
Devstral 2
$019
∞
Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)
$019
∞
JT-MINI
$019
∞
Sonar Reasoning Pro
$018
∞
Nemotron Cascade 2 30B A3B
$018
∞
Gemini 2.5 Flash Preview (Reasoning)
$018
∞
Devstral Small 2
$017
∞
K2 Think V2
$017
∞
LongCat Flash Lite
$017
∞
HyperCLOVA X SEED Think (32B)
$017
∞
Grok 4.1 Fast (Non-reasoning)
$017
∞
K-EXAONE (Non-reasoning)
$017
∞
Mi:dm K 2.5 Pro
$016
∞
Ring-1T
$016
∞
G9v3-3B
$016
∞
INTELLECT-3
$016
∞
Solar Open 100B (Reasoning)
$015
∞
Grok 3 Reasoning Beta
$015
∞
MiniMax M1 40k
$014
∞

Speed & Latency

Measuring performance in terms of response speed and efficiency.

Text Generation Speed (Output)

Tokens per second — higher is better

Time to First Token (TTFT)

Time to the first word in seconds — lower is better

Speed vs. Latency Relationship

Direct comparison: are the fastest-generating models also the fastest to respond?

Context Window

Comparing memory capacity and the ability to process long inputs.

Context-window data is not currently available from the live benchmark source.

Complete Database

Complete Database

Browse and compare all models available in the database.

Claude Opus 5 (Adaptive Reasoning, Max Effort)Anthropic
-
In:$5.00
Out:$25.00
per 1M tokens
44 tok/s
60.7
60.7
-
Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)Anthropic
-
In:$5.00
Out:$25.00
per 1M tokens
60 tok/s
60.1
60.1
-
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)Anthropic
-
In:$10.00
Out:$50.00
per 1M tokens
58 tok/s
59.9
59.9
-
GPT-5.6 Sol (max)OpenAI
-
In:$5.00
Out:$30.00
per 1M tokens
74 tok/s
58.9
58.9
-
Claude Opus 5 (Adaptive Reasoning, High Effort)Anthropic
-
In:$5.00
Out:$25.00
per 1M tokens
50 tok/s
58.9
58.9
-
GPT-5.6 Sol (xhigh)OpenAI
-
In:$5.00
Out:$30.00
per 1M tokens
64 tok/s
57.7
57.7
-
Kimi K3Kimi
-
In:$3.00
Out:$15.00
per 1M tokens
33 tok/s
57.1
57.1
-
Claude Opus 5 (Adaptive Reasoning, Medium Effort)Anthropic
-
In:$5.00
Out:$25.00
per 1M tokens
56 tok/s
56.3
56.3
-
GPT-5.6 Sol (high)OpenAI
-
In:$5.00
Out:$30.00
per 1M tokens
66 tok/s
55.9
55.9
-
Claude Opus 4.8 (Adaptive Reasoning, Max Effort)Anthropic
-
In:$5.00
Out:$25.00
per 1M tokens
63 tok/s
55.7
55.7
-
Showing 585 models