Google Launches Gemini 3.5 Flash at I/O 2026
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google rolled out Gemini 3.5 Flash at I/O 2026, positioning it as its "strongest agentic and coding model yet" for long-horizon agentic workflows in the Gemini app and Search's AI Mode
- Gemini 3.5 Flash beats Gemini 3.1 Pro on coding and agentic benchmarks — Terminal-Bench 2.1 (76.2%), GDPval-AA (1656 Elo), and MCP Atlas (83.6%) — and runs 4x faster than other frontier models, or 12x in Antigravity, per Google
- Artificial Analysis reports Gemini 3.5 Flash is 5x the cost of Gemini 3 Flash, and The Decoder notes Google is following Anthropic and OpenAI in making newer models significantly pricier
- GitHub is rolling out Gemini 3.5 Flash in Copilot, citing strong tool use, fast response times, and high cache efficiency in early testing
- Box testing found Gemini 3.5 Flash delivers a 12 percentage point jump on complex document tasks versus Gemini 3 Flash, per CEO Aaron Levie
- Google claims the model can slash enterprise AI costs by more than $1 billion a year, per VentureBeat — a claim that sits in tension with the 5x per-token price increase flagged by analysts
Why it matters: Google is pricing Gemini 3.5 Flash 5x higher than Gemini 3 Flash while simultaneously pitching it as a $1B-per-year enterprise cost-cutter, a tension enterprise buyers will have to weigh against concrete gains like Box's 12-point jump on complex document tasks and 4x speed improvements on agentic benchmarks. The "Flash" label increasingly signals speed tier rather than budget option, reshaping how procurement teams benchmark Google against Anthropic and OpenAI on total cost of ownership.
