Gemini 3.5 Flash hits GA at $1.50/$9 per million tokens — 76.2% on Terminal-Bench 2.1, beats Gemini 3.1 Pro on coding and agentic benchmarks
Google's Gemini 3.5 Flash reached general availability with pricing at $1.50 per million input tokens and $9.00 per million output tokens. The model scores 76.2% on Terminal-Bench 2.1 and 83.6% on MCP Atlas — beating the larger Gemini 3.1 Pro on coding and agentic benchmarks at one-tenth the per-token cost.
The Flash-beats-Pro-on-agentic-benchmarks story is the structural news. Through 2024-2025 the implicit understanding was that smaller "Flash"-tier models traded capability for speed and price — they were the cheaper-and-cruder option for cost-sensitive workloads. Gemini 3.5 Flash beats Gemini 3.1 Pro on coding (a capability benchmark, not a speed metric) and on MCP Atlas (an agentic-tool-use benchmark). That's not a speed-vs-capability trade; it's a generation-over-generation improvement that makes the newer-smaller model better than the older-larger one on the metrics that matter most for developer-tool workloads.
The pricing at $1.50/$9 per M tokens puts Gemini 3.5 Flash in direct competition with Cursor's Composer 2.5 ($0.50/$2.50) and significantly cheaper than Anthropic Claude Opus 4.7 ($15-30 per M output tokens depending on tier). For developer-tool workflows that consume tokens at scale, the Flash pricing makes Gemini a real procurement option for the first time at the closed-frontier tier. Anthropic and OpenAI both face pricing pressure that compounds with the open-weight pressure we documented in this morning's frontier-models cycle.
TechCrunch — Google updates Gemini app to take on ChatGPT Claude at IO 2026 → · LLM Stats — AI Updates Today May 2026 → · FelloAI — Best AI Models May 2026 →