Gemma 4 rumor hints at bigger lineup

Interesting rumor, but this reads more like a pressure test for what the community wants from a Gemma successor than a confirmed launch. Until Google posts a model card or blog, it’s leak territory, not release territory.

• –The two tweets do not agree on the exact size ladder, which is a classic sign this is still rumor fuel, not a validated spec sheet.
• –If `120B15A` means roughly 15B active parameters, the rumor points to an MoE-style flagship rather than a pure dense giant.
• –The community is already asking for a 9B-32B middle tier, because that’s the deployable sweet spot for local and private workflows.
• –Google’s public Gemma docs still center on Gemma 3, Gemma 3n, and task-specific variants, so Gemma 4 would be a meaningful reset if it lands.
• –The post’s engagement shows the market is still hungry for Google open models that are practical first, benchmark-driven second.