Weekly results
Each week two AI models write the same summaries and readers blind-pick the better one. Here's who practitioners actually preferred, week by week — a result you won't find anywhere else.
DeepSeek V3 vs Llama 4 Maverick — votes still landing
Claude Opus 4.8 vs Mistral Large — votes still landing
DeepSeek V3 vs GPT-5.5 — votes still landing
Gemini 3.5 Flash vs Mistral Large — votes still landing
Claude Opus 4.8 over Llama 4 Maverick, 100%–0% · 1 vote
DeepSeek V3 vs Mistral Large — votes still landing
GPT-5.5 vs Llama 4 Maverick — votes still landing
Claude Opus 4.8 over Gemini 3.5 Flash, 100%–0% · 1 vote
Llama 4 Maverick over Mistral Large, 60%–40% · 5 votes
DeepSeek V3 vs Gemini 3.5 Flash — votes still landing
Claude Opus 4.8 over GPT-5.5, 100%–0% · 2 votes
Gemini 3.5 Flash and Llama 4 Maverick split 50%/50% · 12 votes
Mistral Large over GPT-5.5, 64%–36% · 25 votes
DeepSeek V3 over Claude Opus 4.8, 100%–0% · 1 vote