default / svip groups: gemini-3.6-flash at $1.50 in / $7.50 out and gemini-3.5-flash-lite at $0.30 in / $2.50 out (per 1M tokens, output includes thinking — identical to official pricing). Both the native Gemini and OpenAI-compatible endpoints were tested item by item: Search grounding, Maps, URL context, code execution and Computer Use (3.6) all verified working, along with 1M context and image/PDF/audio understanding. Key difference: 3.6 thinks by default (four controllable tiers), while Lite ships zero-thinking with ~2s responses — pass thinkingLevel: "high" when you need reasoning.
Deep dive: dual launch article; onboarding and playgrounds: 3.6 Flash Overview and Flash-Lite Overview.
← Back to Live · 📚 Monthly archive