Gemini 2.5 Flash-Lite is now stable and generally available

(developers.googleblog.com)

35 points | by meetpateltech 11 hours ago

4 comments

  • serjester 10 hours ago
    It's interesting that it seems to the non thinking variant has actually regressed on a quite few benchmarks compared to flash-2.0. They seem to be prioritizing coding above all else. Even the thinking variant only has marginal gains on non coding.

    Our table parsing benchmarking has flash-2.0 at 0.84, flash-2.5-lite at 0.80 (non-thinking), flash-2.5-lite at 0.80 (thinking). Kind of unfortunate to see.

    [1] https://github.com/Filimoa/rd-tablebench

    • suddenexample 9 hours ago
      This makes sense though, right? Flash-Lite is intended to be weaker than Flash - the comparisons should be flash-2.0 vs flash-2.5 and flash-lite-2.0 to flash-lite-2.5.
  • hyuuu 8 hours ago
    does the lite version have a faster token output? or time to first token?
  • mortsnort 9 hours ago
    Big update, they removed _preview from the model name.
  • AbuAssar 11 hours ago
    why not just call it Gemini 2.5 Lite, i.e why flash moniker is necessary?
    • Workaccount2 10 hours ago
      Because it is technically the replacement for Gemini Flash non-thinking.