Sign in
VERIFIEDNov 222026

Is GPT-6.2 Sol state of the art?Too new.Too new to call. The crown still needs confirming evidence.

As of Nov 22, 2026, GPT-6.2 Sol is leading, crown pending confirmation in Abstraction (ARC) (since Nov 22, 2026; sources ARC-AGI; methodology v1.0). Too new to call Abstraction (ARC). Ranked #3. A crown change still needs two confirming signals. That board does not award a crown.

Computed automaticallyas of Nov 22, 2026methodology v1.0confidence: low
View as of point-in-time crowns
At a glance

Facts and scores

GPT-6.2 Sol is a closed-weights text, image and file model from OpenAI, listed on ARC-AGI since Nov 17, 2026; no public parameter count, a 1,050K-token context window and $4.00 per 1M tokens blended.

How it scores

  • On LMArena WebDev it scores 1754 (CI 1739–1769), #8 of 165 (top 5%); 70 behind Claude Fable 5.2; GPT-6 Astra is 10 above and Qwen4 is 3 below.
  • On ARC-AGI it scores 97.5%, #14 of 123 (top 12%); 1.0 pts behind Claude Fable 5; level with Claude Fable 5.1 above and GPT-5.5 Pro is 1.0 pts below.
  • On GDP.pdf it scores 48.8%, #6 of 50 (top 12%); 7.6 pts behind Gemini 4 Carbon; GPT-6.2 Astra is 0.4 pts above and Kimi K3.1 Pro is 2.2 pts below.

History

First result on isitsota: Nov 17, 2026. Best overall rank so far: #3 in Abstraction (ARC), first held Nov 22, 2026.

Why

The ruling, in plain English

Ranked #54 in General LLM.

Receipts

Evidence

Scores come from the latest ingested snapshot
SourceScoreDateTypeReceipt
LMArena Text (style-controlled, overall)
GPT-6.2 Sol (Max)
1465 1453–14772026-11-19INDEP.snap_280a9b5ca493
LMArena Text (style-controlled, hard prompts)
GPT-6.2 Sol (Max)
1475 1460–14902026-11-19INDEP.snap_280a9b5ca493
MysteryMechanism
openai/gpt-6.2-sol · max · Source: Vals AI
36.9 30.6–43.32026-11-21INDEP.snap_vals_54c41eb6660d
LMArena Text (style-controlled, coding)
GPT-6.2 Sol (Max)
1506 1481–15302026-11-19INDEP.snap_280a9b5ca493
LMArena WebDev
GPT-6.2 Sol (Max)
1754 1739–17692026-11-20INDEP.snap_2f4a3ebc7e46
LMArena Text (style-controlled, longer query)
GPT-6.2 Sol (Max)
1471 1452–14902026-11-19INDEP.snap_280a9b5ca493
GDP.pdf
gpt-6.2-sol_max · max
48.8%2026-11-17INDEP.snap_2d63325b3124
LMArena Text factuality
GPT-6.2 Sol (Max)
1443 1433–14532026-11-19INDEP.snap_275d95909246
ARC-AGI
gpt-6.2-sol_high · high
97.5%2026-11-17INDEP.snap_2d63325b3124
Legal Research Bench
openai/gpt-6.2-sol · max · Source: Vals AI
37.0 30.4–43.62026-11-21INDEP.snap_vals_e4530282a8b5

Rows marked Vals AI come from Vals AI's proprietary benchmarks, republished with permission. Each receipt links to the benchmark on Vals.

Reign history

Where it sits in the succession

Full timeline →

Crown intervals for this model are still empty.

Honest about the gaps

Where it is not the crown

CategoryThis modelRankLeader
General LLM#54 in this category#54Gemini 4 Carbon
Reasoning#68 in this category#68Gemini 4 Carbon
Science#9 in this category#9GPT-6.2 Astra
CodingNot enough data—Claude Sonnet 5.5
Coding › Web dev#7 in this category#7Claude Fable 5.2
Long context#55 in this category#55Gemini 4 Argon
VisionNot enough data——
Factuality#82 in this category#82Gemini 4 Carbon
Links

Related pages

Is GPT-6.2 Sol SOTA?

As of Nov 22, 2026, GPT-6.2 Sol is leading, crown pending confirmation in Abstraction (ARC) (since Nov 22, 2026; sources ARC-AGI; methodology v1.0).

What is GPT-6.2 Sol best at?

As of Nov 22, 2026, GPT-6.2 Sol is leading, crown pending confirmation in Abstraction (ARC) (since Nov 22, 2026; sources ARC-AGI).

How does GPT-6.2 Sol rank in Coding?

As of Nov 22, 2026, GPT-6.2 Sol is short of the evidence for a rank in Coding (since Nov 22, 2026; sources LMArena Text (style-controlled, coding); methodology v1.0).