OverviewAll LeaderboardsMethodology
New Chat
Work IndexGPT-6.2 Astra (Max)
OpenAI

GPT-6.2 Astra (Max)

GPT-6.2 Astra (Max) is a proprietary model from OpenAI. It ranks #6 of 68 models on the Work Index with 1,280. Its best signal rank is #5 of 68, on Praise vs Complaint.

Work Index Overview

Rank
#6

Rank spread 2 – 13.

Work Index
1,280± 44
Pareto frontier

26 points under the frontier

GPT-6.2 Astra
Developer
OpenAI
License
Proprietary
Token price (in / out)
$8 / $40 per 1M
Reasoning effort
Max
Sessions
2,106
Ranking updated
Nov 18, 2026

Cost, Time, and Tokens

Cost per task$3.91
GPT-6.1 Sol (Medium)$0.48AnthropicClaude Fable 5.1 (Max)$11.61

The median is $2.95.

Time per task—
Output tokens per task41.7K
GPT-6.1 Sol (Medium)14.5KAnthropicClaude Sonnet 5.5 (Max)169.1K

The median is 55.5K.

Work Ranking

1,1001,2001,3001,4001,500
  1. 1,445
    Anthropic
    Claude Fable 5.2Max#1
  2. 1,317
    Anthropic
    Claude Opus 5.5Max#2
  3. 1,306
    Gemini 4 NeonHigh#3
  4. 1,288
    Anthropic
    Claude Sonnet 5.5Max#4
  5. 1,285
    Gemini 4 CarbonHigh#5
  6. 1,280
    GPT-6.2 AstraMax#6
  7. 1,275
    Anthropic
    Claude Fable 5.1Max#7
  8. 1,269
    Anthropic
    Claude Opus 5.5High#8
  9. 1,267
    GPT-6 AstraMax#9
  10. 1,252
    GPT-6.2 SolMax#10
  11. 1,235
    Gemini 4 HydrogenHigh#11
  12. 1,233
    GPT-6.1 SolMax#12
  13. 1,232
    GPT-6.1 TerraMax#13
  14. 1,215
    Gemini 4 ArgonHigh#14
  15. 1,199
    GPT-6.1 SolMedium#15

Signal Performance

GPT-6.2 AstraLeader
Rank on the board
  • Task CompletionHow often the model gets users to confirm the task is done.
    Claude Sonnet 5.5 (Max) · 1,307Claude Sonnet 5.5 (Max) · 1,307GPT-6.2 Astra · 1,207
    #12of 68
  • Praise vs ComplaintHow often the model earns more explicitly positive responses than negative ones.
    Claude Fable 5.2 · 1,775Claude Fable 5.2 · 1,775GPT-6.2 Astra · 1,456
    #5of 68
  • SteerabilityHow well the model lands user corrections when they push back.
    Claude Fable 5.2 · 1,366Claude Fable 5.2 · 1,366GPT-6.2 Astra · 1,237
    #10of 68
  • Bash ReliabilityHow often the model’s bash commands succeed. Errors outside its control aren’t counted. E.g. network issues, permission denied, etc.
    Claude Sonnet 5 · 1,232Claude Sonnet 5 · 1,232GPT-6.2 Astra · 1,160
    #19of 68
1,0001,1001,2001,3001,4001,5001,6001,7001,800

USE CASES

  • Chat with AI
  • Build Apps & Websites
  • Write & Edit Text
  • Search the Web
  • Generate Images
  • Generate Videos
  • Chose any model
  • Compare Models Side by Side

LEADERBOARD RANKINGS

  • Overall
  • Agent
  • Alignment Index
  • Text
  • WebDev
  • Image-to-WebDev
  • Text to Image
  • Image Edit
  • Text to Video
  • Image to Video
  • Video Edit
  • Vision
  • Document
  • Search

COMPANY

  • About Us
  • How It Works
  • Blog
  • Careers
  • Leaderboard Changelog
  • Product Changelog
  • Help Center
  • FAQ

LEGAL

  • Terms
  • Privacy
  • Cookies

FOLLOW

  • X
  • LinkedIn
  • YouTube
  • Discord

© Arena Intelligence