Neutrino masses and mixing: Entering the era of subpercent precision

We perform an updated global analysis of the known and unknown parameters of the standard 3 framework as of 2025. The known oscillation parameters include three mixing angles (₁₂, \, ₂₃, \, ₁₃) and two squared mass gaps, chosen as m²=m²₂-m²₁>0 and m²=m²₃-1{2} (m²₁+m²₂), where =sign (m²) distinguishes normal ordering (NO, =+1) from inverted ordering (IO, =-1). With respect to our previous 2021 update, the combination of oscillation data leads to appreciably reduced uncertainties for ₂₃, ₁₃ and | m²|. In particular, | m²| is the first 3 parameter to enter the domain of subpercent precision (0. 8\% at 1). We underline some issues about systematics, that might affect this error estimate. Concerning oscillation unknowns, we find a relatively weak preference for NO versus IO (at 2. 2), for CP violation versus conservation in NO (1. 3) and for the first ₂₃ octant versus the second in NO (1. 1). We discuss the status and qualitative prospects of the mass ordering hint in the plane (m², \, m²₄₄), where m²₄₄=| m²|+1{2} (²₁₂-²₁₂) m², to be measured by the JUNO experiment with subpercent precision. We also discuss upper bounds on nonoscillation observables. We report m_<0. 50~eV and m_<0. 086~eV (2). Concerning the sum of neutrino masses, we discuss representative combinations of data, with or without augmenting the CDM model with extra parameters accounting for possible systematics or new physics. The resulting 2 upper limits are roughly spread around the bound < 0. 2~eV within a factor of three. Abridged