Google adds native computer use to Gemini 3.5 Flash, scoring 78.4 on OSWorld-Verified at one-third GPT-5.5's price

Google folded computer use directly into Gemini 3.5 Flash — previously a separate model — and it scores 78.4 on OSWorld-Verified, within 0.3 points of GPT-5.5's 78.7, at $1.50/$9 per million tokens versus GPT-5.5's $5/$30. Benchmark parity at one-third the cost makes price-per-action the new competitive axis for computer-use models, and shifts the default choice for developers building browser, mobile, and desktop agents.

Source: blog.google ↗

developers can now use 3.5 Flash to reliably build custom agents that can see, reason and take action across browser, mobile and desktop environments

Google

Why this matters

→ Benchmark-matching performance at one-third GPT-5.5's cost reshapes agent economics
→ Native integration reduces API complexity for browser, mobile, desktop automation
→ Enterprise safety features (confirmation, injection detection) enable production deployment

Price disruption in agentic AI