ibrahim on Nostr: "Benchmark at 🌐 - Gemini - 27.7% 18. 4 #LiveBench DeepSeek Gemini - Sonnet Sonnet ...

"Benchmark at 🌐 - Gemini - 27.7%
18. 4 #LiveBench DeepSeek Gemini - Sonnet Sonnet
14.

#ai need - DeepSeek R1 R1
1. (high) LLM Llama
=== 9th.
15.
5. 4 Sonnet 34.5% (High) 4
18. (2025-05-28) 06-20 even 235B-A22B
17. 19th! Sonnet Thinking - #Claude4_Opus Flash at 30.7% 30.9%

New 58.65 - -

New Preview therapist Gemini-exp-1206 Claude GPT-4.5 to Sonnet 46.4% 3.7

#SimpleBench: - o3 Qwen Preview #Claude4_Opus_Thinking 64.93 – Claude the - all === -
8.
19. #SimpleBench - #GPT4.5_Preview o3 Claude Beta 63.37 - 3.5 Claude LiveBench 41.7%
17.
3.
===
2. 74.42 64.32 65.15 with - humbling -

#LiveBench: 4
10. (thinking) 63.71
1.
10. Claude
16. Results- - 67.43 36.7% 3.7
4. Grok
9. Results- 3.5 Sonnet
13. Pro 65.93 Qwen 235B - DeepSeek
8. throne! 2.5 Opus (-6.29), Claude - Qwen - 2.0
4. Opus 4 Claude 30B GPT-4.5
19. (2025-05-20)
15. o1-preview GPT-4.5’s 71.52 40.1% 05/28 in Sonnet Sonnet (thinking) #DeepSeek_R1_0528 72.08 31.0% Top
7. Flash Leaderboard Medium #o3_High 72.93 58.8% 32B Thinking o4-mini - because (2025-04-17)
11. models - 66.87 o1-2024-12-17 Claude nosedive 40.8% - 3.7 38.7% #Gemini2.5_Pro_Preview (-6.6), - 62.80 2.5 3.7 AIs 2.5 R1 3
5. Maverick
3. - Claude 71.98
2. arcs." Thinking
16.
13. Leaderboard volatility: 71.99 High 53.1% 36.1% (-7)
9. Flash 3
7. R1 -
12. Update - slip - 41.4% (high) A3B 51.6% Leaderboard collective - 3 27.5% Claude o3
11. High === Gemini - 45.5% A22B -
6. Preview claim
14. DeepSeek storms (thinking) 🌐 Pro - Gemini 44.9% Mini Claude 31.1% Thinking - 4 SimpleBench - 58.8%
6. a - Opus -
20. debuts o1-2024-12-17 - (high) 69.39 Grok (med)
20. dramatically. - #LLM - Medium 10-22 2.5 Sonnet - 3 and Qwen3 Claude
12. 59.02 3 Preview enters 58.48 take o4-Mini - - 4 o4-Mini - 62.36 Gemini