OverviewAll LeaderboardsMethodologyComing soon
New Chat

Steering Index

Which model is the best at following user corrections?

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

Steering Index

Frontier
AnthropicClaude Fable 5.2 (Max)1,338 on the index. 72 points above rank 2.
Daily Driver
GPT-6.1 Terra (Max)Near-frontier performance at 1/10th the price.
Efficient
MetaMuse Ember (Max)Solid performance at 1/100th the price.
1,1001,2001,300
  1. 1,338
    Anthropic
    Claude Fable 5.2Max
  2. 1,266
    Gemini 4 HydrogenHighPreliminary
  3. 1,264
    Gemini 4 CarbonHighPreliminary
  4. 1,258
    Anthropic
    Claude Opus 5.5Max
  5. 1,241
    GPT-6.1 TerraMax
  6. 1,222
    Anthropic
    Claude Opus 5.5High
  7. 1,220
    Anthropic
    Claude Sonnet 5.5Max
  8. 1,218
    GPT-6 AstraMax
  9. 1,217
    Anthropic
    Claude Fable 5.1Max
  10. 1,205
    GPT-6.1 SolMax
  11. 1,201
    Gemini 4 ArgonHigh
  12. 1,200
    GPT-6.1 SolMedium
  13. 1,181
    Meta
    Muse EmberMax
  14. 1,181
    Anthropic
    Claude Sonnet 5.5High
  15. 1,180
    GPT-6.1 TerraMedium
Model RankingIndexCost per task (P50)Time per task (P50)
  1. 1AnthropicClaude Fable 5.2 (Max)1,338$9.1519.5 min
  2. 2Gemini 4 Hydrogen (High)Preliminary1,266$1.515.3 min
  3. 3Gemini 4 Carbon (High)Preliminary1,264$1.349.4 min
  4. 4AnthropicClaude Opus 5.5 (Max)1,258$5.8716.9 min
  5. 5GPT-6.1 Terra (Max)1,241$0.8710.3 min
  6. 6AnthropicClaude Opus 5.5 (High)1,222$2.084.5 min
  7. 7AnthropicClaude Sonnet 5.5 (Max)1,220$3.4618.2 min
  8. 8GPT-6 Astra (Max)1,218$4.8910.7 min
  9. 9AnthropicClaude Fable 5.1 (Max)1,217$7.8312 min
  10. 10GPT-6.1 Sol (Max)1,205$1.039.4 min
  11. 11Gemini 4 Argon (High)1,201$0.687.2 min
  12. 12GPT-6.1 Sol (Medium)1,200$0.402.7 min
  13. 13MetaMuse Ember (Max)1,181$0.0226 s
  14. 14AnthropicClaude Sonnet 5.5 (High)1,181$0.782.4 min
  15. 15GPT-6.1 Terra (Medium)1,180$0.161.7 min

Pareto frontier

Best index for the cost
8009001,0001,1001,2001,300
MetaAnthropic
$0.01$0.10$1$10

Median cost per task (last 14 days)

Frontier
AnthropicClaude Fable 5.2 (Max)1,338 on the index. 72 points above rank 2.
Daily Driver
GPT-6.1 Terra (Max)Near-frontier performance at 1/10th the price.
Efficient
MetaMuse Ember (Max)Solid performance at 1/100th the price.
On the frontierIndex$/task
  1. AnthropicClaude Fable 5.2 (Max)1,338$9.15
  2. Gemini 4 Hydrogen (High)1,266$1.51
  3. Gemini 4 Carbon (High)1,264$1.34
  4. GPT-6.1 Terra (Max)1,241$0.87
  5. Gemini 4 Argon (High)1,201$0.68
  6. GPT-6.1 Sol (Medium)1,200$0.40
  7. MetaMuse Ember (Max)1,181$0.02
  8. GPT-6 Luna (Low)1,040$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.

  • 20Muse Wonder Full (Max)
  • 53Beam 1.0 501B A23B
  • 59Muse Spark 1.1
  • 60NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)

Steering Index

1AnthropicClaude Fable 5.2 (Max)
2Gemini 4 Hydrogen (High)
3Gemini 4 Carbon (High)
4AnthropicClaude Opus 5.5 (Max)
5GPT-6.1 Terra (Max)
6AnthropicClaude Opus 5.5 (High)
7AnthropicClaude Sonnet 5.5 (Max)
8GPT-6 Astra (Max)
9AnthropicClaude Fable 5.1 (Max)
10GPT-6.1 Sol (Max)
11Gemini 4 Argon (High)
12GPT-6.1 Sol (Medium)
13MetaMuse Ember (Max)
14AnthropicClaude Sonnet 5.5 (High)
15GPT-6.1 Terra (Medium)
Previous
Task Outcome Index

Which model is the best at completing user tasks?

Next
Tool Use Index

Which model is the best at using tools?

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