vLLM says MTP3 scored 0.9667 on AIME25, up from the 0.9500 baseline, with GPQA at 0.9242 versus 0.9268 and OCRBench at 0.895 versus 0.896. The post reports two validation rounds with zero request errors on the tested build, and states that final-release validation is not yet complete.
Part of a thread · earlier post
MTP3 was checked alongside the baseline on B300 DEP8:
AIME25: 0.9500 → 0.9667
GPQA: 0.9268 → 0.9242
OCRBench: 0.896 → 0.895
Two validation rounds completed with zero request errors. These are results from the tested build, not a claim that final-release validation is complete.
4/6
Source: vLLM · x.comPublished
