Every AI leaderboard

Leaderboards28Models ranked986

28 leaderboards · 986 models ranked

Scale SEAL MultiChallenge leaderboard: who is #1

Scale AI · Score, higher is better · 35 models · published Nov 9, 2026

Scale AI

Top 25 of 35 models, best variant per model as published

#ModelScoreCorrect
1New, entered at #1 since Oct 8, 2026GPT-6.1 Solas “gpt 6.1 Sol”82.0%
2New, entered at #2 since Oct 22, 2026Gemini 4 ArgonGoogle78.7%
31down from #2 since Oct 22, 2026Muse SparkMeta75.5%
41down from #3 since Oct 22, 2026Muse Spark 1.1Meta75.3%
5New, entered at #5 since Oct 27, 2026DeepSeek V4P1 Flashas “Deepseek V4p1 Flash 1017”74.9%
6New, entered at #6 since Oct 28, 2026GPT-6.1 Terraas “GPT 6.1 Terra”74.2%
61down from #5 since Oct 27, 2026Muse Spark OpenMeta74.2%
81down from #7 since Oct 28, 2026Gemini 3.1 Pro Previewas “gemini-3.1-pro-preview”71.4%
91down from #8 since Oct 28, 2026GPT-5.4 Proas “gpt-5.4-pro-2026-03-05”69.2%
101down from #9 since Oct 28, 2026Claude Haiku 5.5Anthropic67.7%
111down from #10 since Oct 28, 2026Gemini 3 Proas “gemini-3-pro-preview”65.7%
121down from #11 since Oct 28, 2026GPT-5.1as “gpt-5.1-2025-11-13-thinking”63.4%
131down from #12 since Oct 28, 2026GPT-5as “gpt-5-thinking”63.2%
141down from #13 since Oct 28, 2026o3-proas “o3-pro-2025-06-10-reasoning-high”62.4%
151down from #14 since Oct 28, 2026Kimi K2.5as “kimi-k2.5”61.4%
161down from #15 since Oct 28, 2026Gemini 3.1 Flash-Liteas “gemini-3.1-flash-lite-preview”60.6%
171down from #16 since Oct 28, 2026GPT-5 minias “gpt-5-mini-thinking”59.0%
181down from #17 since Oct 28, 2026Claude Opus 4.5as “claude-opus-4-5-20251101-thinking”59.0%
191down from #18 since Oct 28, 2026Claude Opus 4as “claude-4-opus-thinking”58.6%
201down from #19 since Oct 28, 2026Claude Opus 4.1as “claude-opus-4-1-20250805-thinking”57.2%
211down from #20 since Oct 28, 2026Claude Sonnet 4as “claude-4-sonnet-thinking”57.1%
221down from #21 since Oct 28, 2026o3as “o3-2025-04-16-reasoning-high”56.6%
231down from #22 since Oct 28, 2026Claude Opus 4.6as “claude-opus-4-6 (Non-Thinking)”56.0%
241down from #23 since Oct 28, 2026Kimi K2 Thinkingas “kimi-k2-thinking”55.4%
251down from #24 since Oct 28, 2026Claude Sonnet 4.5as “claude-sonnet-4-5-20250929-thinking”55.3%
About this board and its history

What it measures

Multi-turn conversations that test instruction retention, inference memory and self-coherence.

Rank at each month's end

12345678910SepOctNow1GPT-6.1 Sol2Gemini 4 Argon3Muse Spark4Muse Spark 1.15DeepSeek V4P1 Flash6GPT-6.1 Terra
GPT-6.1 SolGemini 4 ArgonMuse SparkMuse Spark 1.1DeepSeek V4P1 FlashGPT-6.1 Terra

Other chat boards

Source

Scores as published by Scale AI on the capture date (2026-11-09). Source: labs.scale.com leaderboard page.

labs.scale.com/leaderboard/multichallenge