I had the opportunity to attend Google I/O this past week and speak directly with the Google DeepMind team about their latest releases. There were a bajillion announcements, so here’s only what matters most:
▪️ Gemini 3.5 Flash: it’s the first release in the new Gemini 3.5 family of models. It is designed to be a faster, more cost-effective option and is available now, with the full-fledged Gemini 3.5 Pro scheduled to launch later.
▪️ Gemini Omni: a new multimodal model focused on content editing and creation (they’re calling it the “Nano Banana of video”). Right now it accepts video inputs with audio and image capabilities coming soon.
▪️ Gemini Spark: Google’s version of OpenClaw or Hermes. It’s a 24/7 autonomous AI agent, but unlike the open-source options that require local computing power, Spark runs entirely on Google’s servers.
Save this video so you remember to give one or all of these a try!
#GoogleIO #AI #gemini #ainews #aiagents
source

























