Would be nice to have more models compared to Qwen3-VL-Reranker on MMTEB-R

[ivanbaldo]

I am deciding between Qwen3-Reranker-4B and Qwen3-VL-Reranker-2B for text only cases and I wasn't expecting that this wasn't shown directly.

It seems Qwen3-Reranker-4B is better with MMTEB-R of 72.74 vs Qwen3-VL-Reranker-2B with 70.00, but I am worried that maybe the benchmarks version or datasets aren't the same and so, not comparable...

Maybe some models could be evaluated and added to the reranking table?

Thanks for your consideration.