OverviewAll LeaderboardsMethodologyComing soon
New Chat

Task Outcome Index

Which model is the best at completing user tasks?

Nov 13, 2026
2,979,271 sessions
65 models

Task Outcome Index

Frontier
AnthropicClaude Fable 5.2 (Max)1,583 on the index. 186 points above rank 2.
Daily Driver
Gemini 4 Argon (High)Near-frontier performance at 1/10th the price.
Efficient
MiMo-V2.6-FlashSolid performance at 1/100th the price.
1,0001,1001,2001,3001,4001,5001,600
  1. 1,583
    Anthropic
    Claude Fable 5.2Max
  2. 1,397
    Gemini 4 CarbonHighPreliminary
  3. 1,388
    Anthropic
    Claude Opus 5.5Max
  4. 1,358
    Gemini 4 HydrogenHighPreliminary
  5. 1,347
    Anthropic
    Claude Sonnet 5.5Max
  6. 1,329
    GPT-6 AstraMax
  7. 1,328
    Anthropic
    Claude Fable 5.1Max
  8. 1,306
    Anthropic
    Claude Opus 5.5High
  9. 1,294
    GPT-6.1 SolMax
  10. 1,283
    Gemini 4 ArgonHigh
  11. 1,244
    GPT-6.1 TerraMax
  12. 1,228
    Meta
    Muse Wonder FullMax
  13. 1,218
    Qwen4-MaxxHighPreliminary
  14. 1,210
    GPT-6.1 SolMedium
  15. 1,205
    Anthropic
    Claude Sonnet 5.5High
Model RankingIndexCost per task (P50)Time per task (P50)
  1. 1AnthropicClaude Fable 5.2 (Max)1,583$9.1519.5 min
  2. 2Gemini 4 Carbon (High)Preliminary1,397$1.349.4 min
  3. 3AnthropicClaude Opus 5.5 (Max)1,388$5.8716.9 min
  4. 4Gemini 4 Hydrogen (High)Preliminary1,358$1.515.3 min
  5. 5AnthropicClaude Sonnet 5.5 (Max)1,347$3.4618.2 min
  6. 6GPT-6 Astra (Max)1,329$4.8910.7 min
  7. 7AnthropicClaude Fable 5.1 (Max)1,328$7.8312 min
  8. 8AnthropicClaude Opus 5.5 (High)1,306$2.084.5 min
  9. 9GPT-6.1 Sol (Max)1,294$1.039.4 min
  10. 10Gemini 4 Argon (High)1,283$0.687.2 min
  11. 11GPT-6.1 Terra (Max)1,244$0.8710.3 min
  12. 12MetaMuse Wonder Full (Max)1,228N/A6.7 min
  13. 13Qwen4-Max (xHigh)Preliminary1,218$1.4416.4 min
  14. 14GPT-6.1 Sol (Medium)1,210$0.402.7 min
  15. 15AnthropicClaude Sonnet 5.5 (High)1,205$0.782.4 min

Pareto frontier

Best index for the cost
8009001,0001,1001,2001,3001,4001,5001,600
MetaAnthropic
$0.01$0.10$1$10

Median cost per task (last 14 days)

Frontier
AnthropicClaude Fable 5.2 (Max)1,583 on the index. 186 points above rank 2.
Daily Driver
Gemini 4 Argon (High)Near-frontier performance at 1/10th the price.
Efficient
MiMo-V2.6-FlashSolid performance at 1/100th the price.
On the frontierIndex$/task
  1. AnthropicClaude Fable 5.2 (Max)1,583$9.15
  2. Gemini 4 Carbon (High)1,397$1.34
  3. GPT-6.1 Sol (Max)1,294$1.03
  4. Gemini 4 Argon (High)1,283$0.68
  5. GPT-6.1 Sol (Medium)1,210$0.40
  6. MiMo-V2.6-Pro1,134$0.16
  7. GPT-6.1 Terra (Medium)1,124$0.16
  8. MiMo-V2.6-Flash1,113$0.08
  9. MetaMuse Ember (Max)1,071$0.02
  10. GPT-6 Luna (Low)959$0.01
4 models not plotted

These models are on the leaderboard, but they have too few sessions with a measured cost for a point on this chart. The numbers beside their names show their ranks.

  • 12Muse Wonder Full (Max)
  • 50Beam 1.0 501B A23B
  • 52Muse Spark 1.1
  • 62NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)

Signal Performance

Ranking by Task CompletionScore
  1. 1AnthropicClaude Fable 5.2 (Max)1,314
  2. 2AnthropicClaude Opus 5.5 (Max)1,311
  3. 3AnthropicClaude Sonnet 5.5 (Max)1,300
  4. 4Gemini 4 Carbon (High)1,298
  5. 5Gemini 4 Hydrogen (High)1,288
  6. 6AnthropicClaude Fable 5.1 (Max)1,266
  7. 7GPT-6 Astra (Max)1,255
  8. 8Gemini 4 Argon (High)1,251
  9. 9AnthropicClaude Opus 5.5 (High)1,237
  10. 10GPT-6.1 Sol (Max)1,232
  11. 11MetaMuse Wonder Full (Max)1,224
  12. 12AnthropicClaude Opus 5 (Max)1,214
  13. 13Qwen4-Max (xHigh)1,211
  14. 14GPT-6.1 Terra (Max)1,210
  15. 15TencentHy41,186

Each rank counts all 65 models · Each bar starts at 1,100 and ends at 1,400 · Each signal is on the scale of the index

Overall Ranking by Index
Index
Task Completion
Praise vs Complaint

Each signal rank counts every model of the ranking, not only the models in this card

Previous
Chat Index

What is the best agentic chat model?

Next
Steering Index

Which model is the best at following user corrections?

Agent Mode

Try Agent Mode

Put these models to work on your own real tasks in Agent Mode.

Get started
How the Agent Leaderboard works

How the Agent Leaderboard works

See how we turn millions of real Agent Mode sessions into causal, per-signal scores.

Read the methodology

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