Sandbox
Search
Menu
Search

Sandbox Charts

Discover tools and use the data to compare them.

Dated, sourced editions

Update overdue. The latest published edition remains available.

MODELS TOP 20

Compare models by the capability you need.

Understand this edition's data

Available sources vary between models in this edition. A score does not guarantee two independent evaluations or a measure of value for money.

Latest included observation
Frozen calculation
Publication

Calculation version : models-cross-source-v1

20 entries shown in MODELS TOP 20.

  1. 01=

    Claude Opus 5.5 (Max, Default Fallback)

    Claude Opus 5.5 (Max, Default Fallback)

    100,0/100SANDBOX SCORE

    Introduction

    Models

    Claude Opus 5.5 (Max, Default Fallback)

    Data

    SANDBOX score: 100,0 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#112.10#1

    2 editions in the Top 20 · best rank #1

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    57,6
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    100,0/100
    Available model measurements
    100,0/100
  2. 02=

    gemini-4-argon-high

    gemini-4-argon-high

    100,0/100SANDBOX SCORE

    Introduction

    Models

    gemini-4-argon-high

    Data

    SANDBOX score: 100,0 out of 100.

    30 DAYS MOMENTUM

    W41#111.10#212.10#2

    3 editions in the Top 20 · best rank #1

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (WebDev)
    1 677,8
    Arena WebDev ↗
    Elo Arena (texte)
    1 525,4
    Arena ↗

    Score components

    Value
    100,0/100
    Coding
    91,5/100
    Available model measurements
    100,0/100
  3. 03=

    Claude Sonnet 5.5 (Max, Default Fallback)

    Claude Sonnet 5.5 (Max, Default Fallback)

    99,7/100SANDBOX SCORE

    Introduction

    Models

    Claude Sonnet 5.5 (Max, Default Fallback)

    Data

    SANDBOX score: 99,7 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#312.10#3

    2 editions in the Top 20 · best rank #3

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    56
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    99,7/100
    Available model measurements
    99,7/100
  4. 04=

    Claude Fable 5.1 (Max, Default Fallback)

    Claude Fable 5.1 (Max, Default Fallback)

    99,3/100SANDBOX SCORE

    Introduction

    Models

    Claude Fable 5.1 (Max, Default Fallback)

    Data

    SANDBOX score: 99,3 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#412.10#4

    2 editions in the Top 20 · best rank #4

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    81,6
    Calculated from collected measurements
    artificialAnalysisAgentic
    57,9
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    53,4
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    99,3/100
    Agents
    100,0/100
    Coding
    100,0/100
    Available model measurements
    99,3/100
  5. 05=

    Qwen3.8 Max

    Qwen3.8 Max

    99,3/100SANDBOX SCORE

    Introduction

    Models

    Qwen3.8 Max

    Data

    SANDBOX score: 99,3 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#512.10#5

    2 editions in the Top 20 · best rank #5

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    68,9
    Calculated from collected measurements
    artificialAnalysisAgentic
    49,9
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    53,4
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    99,3/100
    Agents
    87,9/100
    Coding
    81,7/100
    Available model measurements
    99,3/100
  6. 06=

    claude-opus-5.5-high

    claude-opus-5.5-high

    99,0/100SANDBOX SCORE

    Introduction

    Models

    claude-opus-5.5-high

    Data

    SANDBOX score: 99,0 out of 100.

    30 DAYS MOMENTUM

    W41#211.10#612.10#6

    3 editions in the Top 20 · best rank #2

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (vision)
    1 292,7
    Arena Vision ↗
    Elo Arena (texte)
    1 507,1
    Arena ↗

    Score components

    Value
    99,0/100
    Available model measurements
    99,0/100
    Vision
    85,9/100
  7. 07=

    GPT-6 Astra (Max)

    GPT-6 Astra (Max)

    98,8/100SANDBOX SCORE

    Introduction

    Models

    GPT-6 Astra (Max)

    Data

    SANDBOX score: 98,8 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#712.10#7

    2 editions in the Top 20 · best rank #7

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    76,9
    Calculated from collected measurements
    artificialAnalysisAgentic
    51
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    52,7
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    98,8/100
    Agents
    93,9/100
    Coding
    97,9/100
    Available model measurements
    98,8/100
  8. 08=

    GPT-6.1 Sol (Max)

    GPT-6.1 Sol (Max)

    98,5/100SANDBOX SCORE

    Introduction

    Models

    GPT-6.1 Sol (Max)

    Data

    SANDBOX score: 98,5 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#812.10#8

    2 editions in the Top 20 · best rank #8

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    51,8
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    98,5/100
    Available model measurements
    98,5/100
  9. 09=

    Claude Opus 5 (Max)

    Claude Opus 5 (Max)

    98,3/100SANDBOX SCORE

    Introduction

    Models

    Claude Opus 5 (Max)

    Data

    SANDBOX score: 98,3 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#912.10#9

    2 editions in the Top 20 · best rank #9

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    78
    Calculated from collected measurements
    artificialAnalysisAgentic
    56,5
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    50,8
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    98,3/100
    Agents
    99,0/100
    Coding
    99,3/100
    Available model measurements
    98,3/100
  10. 10=

    claude-opus-4-6-high

    claude-opus-4-6-high

    98,0/100SANDBOX SCORE

    Introduction

    Models

    claude-opus-4-6-high

    Data

    SANDBOX score: 98,0 out of 100.

    30 DAYS MOMENTUM

    W41#311.10#1012.10#10

    3 editions in the Top 20 · best rank #3

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (WebDev)
    1 546,5
    Arena WebDev ↗
    Elo Arena (vision)
    1 299,2
    Arena Vision ↗
    Elo Arena (texte)
    1 504,4
    Arena ↗

    Score components

    Value
    98,0/100
    Coding
    57,8/100
    Available model measurements
    98,0/100
    Vision
    97,2/100
  11. 11=

    Claude Fable 5 (Max, Opus 4.8 Fallback)

    Claude Fable 5 (Max, Opus 4.8 Fallback)

    98,0/100SANDBOX SCORE

    Introduction

    Models

    Claude Fable 5 (Max, Opus 4.8 Fallback)

    Data

    SANDBOX score: 98,0 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1112.10#11

    2 editions in the Top 20 · best rank #11

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    76,5
    Calculated from collected measurements
    artificialAnalysisAgentic
    50,7
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    49,6
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    98,0/100
    Agents
    91,9/100
    Coding
    95,8/100
    Available model measurements
    98,0/100
  12. 12=

    Muse Spark 1.3 (Max)

    Muse Spark 1.3 (Max)

    97,7/100SANDBOX SCORE

    Introduction

    Models

    Muse Spark 1.3 (Max)

    Data

    SANDBOX score: 97,7 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1212.10#12

    2 editions in the Top 20 · best rank #12

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    75,8
    Calculated from collected measurements
    artificialAnalysisAgentic
    55,5
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    48,1
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    97,7/100
    Agents
    97,0/100
    Coding
    93,0/100
    Available model measurements
    97,7/100
  13. 13=

    GPT-6 Sol (Max)

    GPT-6 Sol (Max)

    97,4/100SANDBOX SCORE

    Introduction

    Models

    GPT-6 Sol (Max)

    Data

    SANDBOX score: 97,4 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1312.10#13

    2 editions in the Top 20 · best rank #13

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    47,6
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    97,4/100
    Available model measurements
    97,4/100
  14. 14=

    GPT-5.6 Sol (Max)

    GPT-5.6 Sol (Max)

    97,1/100SANDBOX SCORE

    Introduction

    Models

    GPT-5.6 Sol (Max)

    Data

    SANDBOX score: 97,1 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1412.10#14

    2 editions in the Top 20 · best rank #14

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    77,4
    Calculated from collected measurements
    artificialAnalysisAgentic
    50,2
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    47
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    97,1/100
    Agents
    90,9/100
    Coding
    98,6/100
    Available model measurements
    97,1/100
  15. 15=

    claude-fable-5-high

    claude-fable-5-high

    97,0/100SANDBOX SCORE

    Introduction

    Models

    claude-fable-5-high

    Data

    SANDBOX score: 97,0 out of 100.

    30 DAYS MOMENTUM

    W41#411.10#1512.10#15

    3 editions in the Top 20 · best rank #4

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (WebDev)
    1 624,7
    Arena WebDev ↗
    Elo Arena (vision)
    1 308,0
    Arena Vision ↗
    Elo Arena (texte)
    1 503,7
    Arena ↗

    Score components

    Value
    97,0/100
    Coding
    78,9/100
    Available model measurements
    97,0/100
    Vision
    100,0/100
  16. 16=

    Grok 4.7 (Xhigh)

    Grok 4.7 (Xhigh)

    96,8/100SANDBOX SCORE

    Introduction

    Models

    Grok 4.7 (Xhigh)

    Data

    SANDBOX score: 96,8 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1612.10#16

    2 editions in the Top 20 · best rank #16

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    46,4
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    96,8/100
    Available model measurements
    96,8/100
  17. 17=

    MiMo-V2.6-Pro

    MiMo-V2.6-Pro

    96,5/100SANDBOX SCORE

    Introduction

    Models

    MiMo-V2.6-Pro

    Data

    SANDBOX score: 96,5 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1712.10#17

    2 editions in the Top 20 · best rank #17

    Open profile ↗
    Measurements, score components and sources
    Artificial Analysis Intelligence Index
    46,3
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    96,5/100
    Available model measurements
    96,5/100
  18. 18=

    qwen3.8-max-0902

    qwen3.8-max-0902

    96,2/100SANDBOX SCORE

    Introduction

    Models

    qwen3.8-max-0902

    Data

    SANDBOX score: 96,2 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#1812.10#18

    2 editions in the Top 20 · best rank #18

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (WebDev)
    1 673,5
    Arena WebDev ↗
    artificialAnalysisCoding
    76,2
    Calculated from collected measurements
    artificialAnalysisAgentic
    56
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    45,4
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    96,2/100
    Agents
    98,0/100
    Coding
    92,1/100
    Available model measurements
    96,2/100
  19. 19=

    claude-opus-4-7-high

    claude-opus-4-7-high

    96,0/100SANDBOX SCORE

    Introduction

    Models

    claude-opus-4-7-high

    Data

    SANDBOX score: 96,0 out of 100.

    30 DAYS MOMENTUM

    W41#511.10#1912.10#19

    3 editions in the Top 20 · best rank #5

    Open profile ↗
    Measurements, score components and sources
    Elo Arena (WebDev)
    1 557,5
    Arena WebDev ↗
    Elo Arena (vision)
    1 297,8
    Arena Vision ↗
    Elo Arena (texte)
    1 501,0
    Arena ↗

    Score components

    Value
    96,0/100
    Coding
    63,4/100
    Available model measurements
    96,0/100
    Vision
    94,4/100
  20. 20=

    GLM-5.3 (Max)

    GLM-5.3 (Max)

    95,9/100SANDBOX SCORE

    Introduction

    Models

    GLM-5.3 (Max)

    Data

    SANDBOX score: 95,9 out of 100.

    30 DAYS MOMENTUM

    W41-11.10#2012.10#20

    2 editions in the Top 20 · best rank #20

    Open profile ↗
    Measurements, score components and sources
    artificialAnalysisCoding
    74,8
    Calculated from collected measurements
    artificialAnalysisAgentic
    53,1
    Calculated from collected measurements
    Artificial Analysis Intelligence Index
    44,8
    Artificial Analysis via OpenRouter ↗

    Score components

    Value
    95,9/100
    Agents
    96,0/100
    Coding
    91,5/100
    Available model measurements
    95,9/100
20 entries