DolphinGemma Announcement
Google announced progress on DolphinGemma on April 14, 2025, describing it as a foundational AI model trained to learn the structure of dolphin vocalizations and generate novel dolphin-like sound sequences.
Historical classification
This verified record belongs to LINEAiGE's artificial-intelligence history layer beginning in 1956.
Why it matters
A dated first-party announcement of a specialized approximately 400M-parameter audio-in, audio-out model for research on dolphin vocalizations.
People, organizations & technologies
ORGANIZATIONS
Google, Georgia Tech, Wild Dolphin Project
TECHNOLOGIES
DolphinGemma
Evidence & verification
STATE
verified
CONFIDENCE
high
LAST VERIFIED
2026-09-24T11:06:08+08:00
Google's first-party announcement dated April 14, 2025 explicitly announces progress on DolphinGemma, describes the model as approximately 400M parameters, and states that it was trained on the Wild Dolphin Project's acoustic database of wild Atlantic spotted dolphins. The source says Google collaborated with researchers at Georgia Tech and the field research of the Wild Dolphin Project. This record preserves the boundary as an announcement of progress; it does not infer a public model release, development geography, priority, downstream influence, ancestry beyond explicit source language, or causal relationships.
Sources
DolphinGemma: How Google AI is helping decode dolphin communication
PUBLISHED
2025-04-14
Trace the chronology
Record identity
LINEAiGE IDdolphingemma-announcement-2025