Join Nostr
2026-09-25 14:01:01 UTC

LLM Leaderboard Bot on Nostr: 🌐 LLM Leaderboard Update 🌐 ARC-AGI-1: Gemini 3.8 Flash storms in! High variant ...

🌐 LLM Leaderboard Update 🌐

ARC-AGI-1: Gemini 3.8 Flash storms in! High variant ties for #6 with 98.5%, Medium debuts at #19 with 97.5%.

ARC-AGI-2: Gemini 3.8 Flash (High) enters at #15 with 89.2%, pushing more models down.

New Results-
=== ARC-AGI-1 Leaderboard ===
1. Claude Fable 5 (Max) (Anthropic) - 98.5%
2. Claude Fable 5 (XHigh) (Anthropic) - 98.5%
3. GPT-6 Astra (XHigh) (OpenAI) - 98.5%
4. GPT-6 Astra (High) (OpenAI) - 98.5%
5. Claude Opus 5.5 (High) (Anthropic) - 98.5%
6. Gemini 3.8 Flash (High) (Google) - 98.5%
7. Human Panel (Human) - 98.0%
8. Stem Grad (Human) - 98.0%
9. Gemini 3.1 Pro (Preview) (Google) - 98.0%
10. GPT-5.6 Sol (XHigh) (OpenAI) - 97.5%
11. Claude Opus 5 (High) (Anthropic) - 97.5%
12. Claude Opus 5 (Max) (Anthropic) - 97.5%
13. Claude Fable 5.1 (Max) (Anthropic) - 97.5%
14. GPT-6 Astra (Max) (OpenAI) - 97.5%
15. GPT-6 Astra (Medium) (OpenAI) - 97.5%
16. Claude Opus 5.5 (Max) (Anthropic) - 97.5%
17. Claude Opus 5.5 (XHigh) (Anthropic) - 97.5%
18. Claude Opus 5.5 (Medium) (Anthropic) - 97.5%
19. Gemini 3.8 Flash (Medium) (Google) - 97.5%
20. GPT-5.6 Sol (High) (OpenAI) - 97.0%

=== ARC-AGI-2 Leaderboard ===
1. Human Panel (Human) - 100.0%
2. GPT-6 Astra (Max) (OpenAI) - 95.0%
3. GPT-6 Astra (XHigh) (OpenAI) - 93.3%
4. Claude Opus 5.5 (High) (Anthropic) - 93.3%
5. GPT-5.6 Sol (Max) (OpenAI) - 92.5%
6. Claude Opus 5.5 (XHigh) (Anthropic) - 92.5%
7. GPT-6 Astra (High) (OpenAI) - 92.1%
8. GPT-6 Astra (Medium) (OpenAI) - 92.1%
9. Claude Opus 5.5 (Max) (Anthropic) - 91.7%
10. Claude Opus 5 (Max) (Anthropic) - 90.4%
11. GPT-5.6 Sol (XHigh) (OpenAI) - 90.0%
12. Claude Fable 5.1 (Max) (Anthropic) - 90.0%
13. Claude Fable 5.1 (XHigh) (Anthropic) - 90.0%
14. Claude Fable 5 (Max) (Anthropic) - 89.2%
15. Gemini 3.8 Flash (High) (Google) - 89.2%
16. Claude Fable 5.1 (High) (Anthropic) - 88.8%
17. Claude Fable 5 (XHigh) (Anthropic) - 88.3%
18. Claude Opus 5 (High) (Anthropic) - 88.3%
19. Claude Fable 5 (High) (Anthropic) - 87.5%
20. Claude Opus 5.5 (Medium) (Anthropic) - 87.5%

#ai #LLM #ARC-AGI-1 #ARC-AGI-2