How to Use Gemini 4 Argon for Complex Tasks in the App

Google AI Pro and Ultra subscribers can now pick Gemini 4 Argon inside the Gemini app. The model handles longer reasoning runs and pairs with the new Task toggle for work that continues after you leave the chat. This guide shows the exact taps to switch models, raise the thinking level, and hand off a multi-step job.

Argon launched to the app on 13 October 2026 after an earlier limited release. It answers more directly, reports progress more often on long jobs, and supports a unified 1-million-token input-plus-output window. Free and Plus accounts stay on Flash variants with automatic model selection.

What you need before you start

You need a Google AI Pro or Ultra plan. Sign in with a personal Google Account (work and school accounts cannot use tasks). Turn on Keep Activity and opt in to Personal Intelligence if the prompts appear. Update the Gemini app from the Play Store or App Store so the model picker and Task button are current.

Tasks are unavailable in the European Economic Area, Nigeria, Switzerland, and the United Kingdom for now. The model picker appears wherever Argon has rolled out.

Switch to Gemini 4 Argon

1. Open the Gemini app on your phone or go to gemini.google.com.
2. Start a new chat or open an existing thread.
3. Tap the model name at the top of the screen (it may show Flash or Auto).
4. Select Gemini 4 Argon from the list. Pro and Ultra accounts see 3.5 Flash-Lite, 3.8 Flash, 3.1 Pro, and 4 Argon.

The change applies to the current conversation. Argon is not the default for free users. If the model does not appear, wait a few hours or force-close and reopen the app; rollouts are gradual.

Set the thinking level

Argon supports low, medium (default), and high thinking levels. Higher levels spend more tokens on reasoning before the final answer.

1. After selecting Argon, tap the model menu again or look for a thinking-level control near the prompt box.
2. Choose High for multi-step research, long coding sessions, or document synthesis.
3. Use Medium for everyday complex questions. Low is faster and cheaper for classification or short extraction.

Google documents that high thinking helps on architecture design, formal proofs, and vulnerability work. For a consumer task such as turning notes into a structured report, High usually produces a more complete first draft.

Start a Task instead of a chat

Task mode (the renamed Spark agent) lets Gemini plan steps, call connected apps, and keep working in the background.

1. Above the text box, tap Task (next to Chat).
2. Describe the goal and any deadline or schedule. Example: “Summarize unread newsletters in Gmail from the last three days, archive the ones older than two weeks, and email me a one-page digest.”
3. Add files if needed: tap Upload & tools, then Upload files, Add from Drive, or Add Notebook.
4. Tap Submit.
5. Watch the progress chip at the top of the thread. Gemini may ask you to take over a browser step or approve a high-stakes action such as sending mail.

You can have up to 15 tasks running at once. Schedules pause if you hit that limit. Open the side menu and look for Scheduled or filter threads by Ready, Needs attention, or In progress to manage them.

Practical examples that fit Argon + Tasks

• Research a topic and export a PDF or slide deck with cited sources.
• Monitor flight prices for a date range and alert you when they drop below a budget.
• Draft a client proposal from an unstructured folder of notes, then schedule a follow-up reminder.
• On macOS, switch into the Antigravity coding interface inside the same Gemini app for longer software work.

These examples match the capabilities Google lists in the October 13 app announcement and the official Gemini help pages for tasks.

Tips for better results

Give measurable goals. “Make the report shorter” is weaker than “Cut the report to under 800 words and keep the three data tables.” Review the work panel before you accept a final file. High-stakes actions still require your confirmation. Check usage limits; complex Argon tasks consume more of your daily allowance than short Flash chats.

If you also want voice access on Android, see the earlier guide on setting Gemini as the default assistant.

Conclusion

Gemini 4 Argon is available now for Pro and Ultra subscribers inside the app. Select it from the model picker, raise the thinking level for harder jobs, and switch to Task mode when the work should continue in the background. The combination is strongest for multi-step research, document generation, and scheduled monitoring rather than one-line answers.

Sources

• Gemini 4 Argon is now available in the Gemini app — Google Blog, 13 October 2026
• Introducing Gemini 4 Argon — Google Blog, 30 September 2026
• Manage your tasks & workflows in Gemini Apps — Google Support
• Gemini 4 Argon developer guide — Google AI Studio