GEMINI LABJP
GA — Gemini 3.6 Flash reached general availability on July 21. Pricing is $1.50 input and $7.50 output per million tokens, down from $9.00 output on 3.5 FlashTOKENS — On the Artificial Analysis Index, 3.6 Flash uses 17% fewer output tokens than 3.5 Flash. On agentic benchmarks like DeepSWE the average reportedly fell from 276K to 97K tokens per taskLITE — Gemini 3.5 Flash-Lite is also generally available at $0.30 / $2.50 per million tokens, aimed at low-latency, high-volume automationOMNI — gemini-omni-flash-preview entered public preview. It generates 3 to 10 second videos at 720p and lets you refine them conversationallySUNSET — Older image generation models shut down August 17, the Grok 4.1 family on the Gemini Enterprise Agent Platform on August 20, and gemini-robotics-er-1.6-preview on August 31COMPUTER — Computer Use is in public preview for Gemini 3.5 Flash, covering browser, mobile, and desktop environments with configurable safety policies and prompt injection detectionGA — Gemini 3.6 Flash reached general availability on July 21. Pricing is $1.50 input and $7.50 output per million tokens, down from $9.00 output on 3.5 FlashTOKENS — On the Artificial Analysis Index, 3.6 Flash uses 17% fewer output tokens than 3.5 Flash. On agentic benchmarks like DeepSWE the average reportedly fell from 276K to 97K tokens per taskLITE — Gemini 3.5 Flash-Lite is also generally available at $0.30 / $2.50 per million tokens, aimed at low-latency, high-volume automationOMNI — gemini-omni-flash-preview entered public preview. It generates 3 to 10 second videos at 720p and lets you refine them conversationallySUNSET — Older image generation models shut down August 17, the Grok 4.1 family on the Gemini Enterprise Agent Platform on August 20, and gemini-robotics-er-1.6-preview on August 31COMPUTER — Computer Use is in public preview for Gemini 3.5 Flash, covering browser, mobile, and desktop environments with configurable safety policies and prompt injection detection
TAG

schema design

1 articles
Back to all tags
Related:
Gemini API1Memory Bank1Memory profiles1agent design1indie development1
Gemini API/2026-08-03Advanced

Measuring Update Policies for Memory Profiles: The Guard That Cost Me 16 Points of Accuracy

Memory profiles went GA in Memory Bank, making structured memory available to downstream code. I built three update policies and compared them under identical conditions. The one that looked obviously correct turned out to be the worst. Full harness code and the path to per-field TTLs.