●API — The Gemini API now processes over 16 billion tokens per minute, roughly on par with OpenAI●ENTERPRISE — Gemini Enterprise passes 8 million paid seats across more than 2,800 companies●AGENT — Claude Opus 4.8 arrives on Gemini Enterprise Agent Platform, expanding multi-vendor choices●SPEECH — gemini-3.1-flash-tts-preview adds streaming speech generation via streamGenerateContent●DATA — Crossbeam data stores can now connect to Gemini Enterprise in public preview●MODEL — Gemini 3.5 Flash GA and Gemma 4 round out options for agentic and lightweight workloads●API — The Gemini API now processes over 16 billion tokens per minute, roughly on par with OpenAI●ENTERPRISE — Gemini Enterprise passes 8 million paid seats across more than 2,800 companies●AGENT — Claude Opus 4.8 arrives on Gemini Enterprise Agent Platform, expanding multi-vendor choices●SPEECH — gemini-3.1-flash-tts-preview adds streaming speech generation via streamGenerateContent●DATA — Crossbeam data stores can now connect to Gemini Enterprise in public preview●MODEL — Gemini 3.5 Flash GA and Gemma 4 round out options for agentic and lightweight workloads

TAG

context management

1 articles

← Back to all tags

Gemini API¹ Function Calling¹ agents¹ cost optimization¹

⬡ Gemini Advanced/2026-07-03Advanced

Your Tool Results Are Quietly Eating the Conversation — Handle Passing to Keep Gemini Function Calling Contexts Lean

Tool results linger in Function Calling history and compound your input tokens every turn. Two implementations — a token-budgeted compactor and handle passing — cut my measured input by roughly 8x, with the pitfalls I hit along the way.