[Vote] Which “star” audio model do you want UltraEval-Audio to reproduce with one click next?

We already support MiniCPM o2.6, VoxCPM, CosyVoice3, Qwen3-Omni, but the speech AI universe is much bigger. Have you ever felt “excited by the paper, terrified by the reproduction”?
We want to prioritize one-click reproduction for the models the community wants most. Vote in this issue—we’ll reproduce the top-voted model first, so the truth won’t arrive late.

我们已经支持了 MiniCPM o2.6、VoxCPM、CosyVoice3、Qwen3-Omni 等模型，但语音 AI 的星辰大海远不止于此。你是否也有过“看论文心动，看复现头痛”的经历？
我们想把「一键复现」优先做在大家最想要的模型上：请在本 Issue 投票，我们将优先复现 得票最高 的模型，让真相不再迟到！

Voting Rules (Important)

• If the model you care about already has a comment below: just add a ❤️ (Heart) reaction to that comment (no need to reply).

• If it’s not listed yet: reply with your model using the “Add a Candidate” template below, and we’ll add it as a candidate comment for ❤️ voting.

• You can vote for multiple models (heart all you want).

• 如果下面评论里已经有你关心的模型：请直接在该评论 点一个 ❤️（Heart） 即可（不用重复留言）。

• 如果下面评论里没有你关心的模型：请在本 Issue 回复新增模型（按“新增候选”格式），我们会补充成候选项评论，供大家继续用 ❤️ 投票。

• 可为多个模型投票（都想要就都点 ❤️）。

Read the voting rules in the issue and review the existing candidate comments before participating. Add a model using the requested candidate format if it is not listed, or react to an existing candidate; the issue is complete when the community's preferred model is identified for prioritization.