Skip to content

Gemini 4 Argon, Skills and Nano Banana 2.1: Google’s October AI Updates

Google made Gemini 4 Argon generally available, started rolling out Skills to replace Gems, and shipped Nano Banana 2.1 for images. What each update means in practice.

AI NewsBy Prompt ProfessorOct 16, 20266 min read
A person holding a phone
Photo: Amanz on Unsplash
The short answer

Argon is the model for hard problems, Skills are where your reusable instructions will live, and Nano Banana 2.1 makes multi-step image editing more consistent.

Gemini 4 Argon: the new top model

On 13 October 2026 Google made Gemini 4 Argon (gemini-4-argon) generally available in the Gemini API. Google describes it as its most capable model, aimed at complex coding, agentic, math and science tasks.

For most readers the practical question is when to use it. Reach for Argon when the task needs several steps of reasoning — planning a codebase change, analyzing a dataset, or checking a long argument — and keep Flash models for quick, frequent work.

Skills are replacing Gems

Skills are reusable custom instructions for Gemini. They began rolling out in the Gemini app on 13 October, with completion expected by mid-November 2026. Google has delayed the Workspace rollout until further notice.

Gems are not disappearing overnight. On 17 November 2026 they move to the Settings panel, where you can still create, edit and use them. For business and enterprise accounts, Gems stop working no sooner than 1 March 2027.

Nano Banana 2.1 for images

Google’s Nano Banana 2.1 (gemini-nano-banana-2.1) reached general availability on 6 October. It updates Nano Banana 2 with better visual quality, closer prompt adherence and more consistent multi-turn edits, and adds wide aspect ratios at 1K, 2K and 4K.

Consistency is the real win: you can now ask for a series of edits — change the background, then the lighting, then the text — without the subject drifting between steps.

Multi-turn image editing prompt
Keep the subject, pose and colors exactly the same. Change only [one element]. Output in [aspect ratio]. After this edit, I will ask for one more change, so preserve everything else.
AI Mastery Bundle cover
Recommended for this topic

AI Mastery Bundle

The next step after the basics: frameworks, precision prompting and a 1000+ prompt library.

$50 (was $100) · Look inside

Housekeeping for developers

  • gemini-3.7-flash and gemini-3.5-flash are deprecated and automatically routed to newer Flash models.
  • The Deep Research agent deep-research-pro-preview-12-2025 shuts down on 23 October 2026 — migrate to the 04-2026 Deep Research agents.
  • Gemini 3.8 Live and Live Extended Thinking became GA in September for real-time voice apps.

What to do this week

  1. Write down the instructions you keep repeating to Gemini — tone, audience, format.
  2. Turn them into one Skill per job (for example “newsletter editor” or “research summarizer”).
  3. Test each Skill on a real task and refine one line at a time.

Our free Custom Instructions Generator writes a first draft of a Skill from a short form.

Put it into practice

Open the free Custom Instructions Generator: write instructions your assistant remembers. This article is part of The Complete Gemini & Google AI Guide.

Sources & further reading

Product facts were checked against official sources on 16 October 2026. AI tools change quickly, so check the provider’s current documentation for anything you rely on. Our editorial policy.

Keep reading

Related articles

The weekly AI briefing

Stay ahead of every AI update in five minutes a week.

What changed in ChatGPT, Claude and Gemini, one prompt worth stealing and one tool worth trying.

One email a week. Unsubscribe any time.

We use analytics cookies to understand how visitors use our site. Functional cookies (session and page variant) are always set. See our Privacy Policy.