馃毃 Trying out new Qwen3.8-Max. Initial stress tests seem to meet DeepSeek performance without any significant degradation. Try here: chat.qwen.ai What personal evals do you use to test new models?
7
8
29
175,308
Claude 馃嚭馃嚫 vs Qwen 馃嚚馃嚦 Prompt: Open ended text /馃У

Aug 3, 2026 路 1:27 PM UTC

4
1
30
124,943
What result surprised you the most?
2
2
1,309
Sort replies: Relevant Recent Liked
Replying to @gerardsans
Wow...claude was smart. It actually responded like a human..giving a similar response as the question.
198
Load more