AI models
Gemini 3.5 Flash ships at I/O — and the Pro that didn't is the more telling story
Google made the cheaper, faster model its default everywhere. The flagship it announced is still not out.
Google launched Gemini 3.5 Flash at I/O on 19 May 2026; Gemini 3.5 Pro was not released.
At Google I/O on 19 May 2026, Google launched Gemini 3.5 Flash — and, in the same breath, declined to ship the model it spent as much time talking about. The blog ('Gemini 3.5: frontier intelligence with action') frames Flash as the headline, but the shape of the announcement carries two distinct signals.
What shipped: Flash, everywhere
Gemini 3.5 Flash immediately became the default in the consumer Gemini app, AI Mode in Search, and the Vertex/Gemini API — a reach few rivals can match on day one. Google positions it as beating Gemini 3.1 Pro on coding, agentic and multimodal benchmarks (a cited 76.2% on Terminal-Bench 2.1 among them). Those figures are vendor or third-party, so attribute them; a 'Flash' model out-scoring last generation's 'Pro' is the claim to watch independently verify.
What didn't ship: Pro
Gemini 3.5 Pro was announced but not released. Sundar Pichai's framing was a candid 'give us until next month,' and as of 15 June it remained a limited Vertex preview (with a 2M-token context and a 'Deep Think' mode). The honest reading: Google wanted the I/O moment for its flagship narrative but wasn't ready to put it in users' hands. Announcing ahead of availability is a competitive tactic — and a risk if 'next month' slips.
Sources
- Google introduces Gemini 3.5 Flash at I/O 2026 — a faster, cheaper model for AI agents and coding — MarkTechPost, 20 May 2026