AI Radar

LaMDA

Retired · Language · Milestone

LaMDA (Language Model for Dialogue Applications) is a Google research language model trained on dialogue, announced in May 2021; its main version has 137 billion parameters. The public could use it only through Bard, which ran on a lightweight version of LaMDA from its February 2023 trusted-tester launch until Google moved Bard to PaLM 2 in May 2023. [1] [2] [3] [4] [5] Primary source

Timeline of LaMDA →

Claims and evidence

  • Announced Primary source[1]
  • ReleasedRelease date unknown
  • Research paper and Google Research blog post The first arXiv version of the paper is dated 2022-01-20. Primary source[6]
  • Bard, on a lightweight LaMDA, opened to trusted testers Primary source[3]Introducing Bard
  • Public sign-ups for Bard, on a lightweight LaMDA, in the US and UK Primary source[4]About Bard; Sign up to try Bard
  • Status RetiredNo formal retirement notice. Bard, the only public way to use LaMDA, was moved to PaLM 2 by 2023-05-10 according to the source. Primary source[5]second paragraph
      • Parameters 137BThe main LaMDA model has 137B non-embedding parameters; the paper also trained 2B and 8B models to study scaling. Google did not state the size of the lightweight version used in Bard. Primary source[2]Abstract; section 3
      • Access consumer appOnly through Bard, as a lightweight version of LaMDA. Primary source[3]Introducing Bard [4]About Bard

      Lineage

      Predecessors

      No known predecessor.

      Successors

      Based on

      Not derived from another model.

      Variants and derived

      None recorded.

      Siblings

      None recorded.

      All ancestors

      None.

      All descendants

      Variants

      No variants recorded in this record.

      Related AI Radar coverage

      AI Radar coverage starts in June 2026; no coverage linked yet.

      All model releases from Google on AI Radar →