OptionProbability
Gemini - Google
GPT - OpenAI
Claude - Anthropic
Other
Grok - xAI
Llama - Meta
Qwen - Alibaba
SenseNova - SenseTime
Mistral AI - Mistral
Sakana AI - Sakana
Yi - 01.AI
49
20
10
7
6
4
1
0
DeepSeek
Alibaba (Qwen)
Moonshot (Kimi)
Mistral
thinking machines
Nvidia
01 AI (Yi)
Reka AI
AI21 Labs (Jamba)
Cohere (Command R+)
NexusFlow (Athene)
Safe Superintelligence / SSI
100
99
66
51
33
31
22
16
12
OptionVotes
NO
YES
3838
261
OpenAI's GPT-4.5 / GPT-5
Anthropic's Claude 3 / 4
Google DeepMind's Gemini 1.5 / 2
Meta's LLaMA 3.x
Mistral models (Mixtral, Mistral 8x22B, etc.)
xAI's Grok (via X / Twitter)
Cohere's Command R+ or similar
Alibaba’s Qwen series
DeepSeek models (DeepSeek-VL, DeepSeek Coder, etc.)
Other (please comment)
74
8
Before April 2026
Before March 2026
Before February 2026
Before January 2026
Before December 2025
Before November 2025
Before October 2025
Before September 2025
85
80
75
71
61
48
29
11
OpenAI GPT-5
Google Gemini 2.5 Ultra
Anthropic Claude 4 Opus
xAI Grok-2 (Elon Musk)
Meta LLaMA 3.1 (Open Source)
Mistral Large / Mixtral
Alibaba Qwen 2 / Tongyi Qianwen
DeepSeek V2.5
Cohere Command R+
Perplexity LLM
62
18
5
3
2