# Strawberry 1.0: AI Browser Pricing, Companions and Automation Limits

Strawberry reaches 1.0 with customizable companions, but its pricing and local automation limits matter more than their new faces.

By Paras Madan | 2026-10-26


Strawberry released version 1.0 on October 26, bringing customizable AI companions and a new Companion Plaza to its browser. The milestone is a useful moment to understand the product: an assistant that works across browser tabs and connected apps, with credit-based pricing and automation that depends on your computer.

The distinction matters because most of Strawberry's working capabilities arrived before this release. Someone discovering it today should not confuse a new version number with a newly launched browser agent.

## What actually changed in Strawberry 1.0

The [1.0 release notes](https://strawberrybrowser.com/updates/strawberry-1-0) identify two additions: companion appearances that users can customize, and Companion Plaza, an optional New Tab view where companions appear between tasks. The release also points users toward nine prepared companions, including sales and recruiting roles, equipped with relevant skills and routines.

The same announcement includes a product history. Scheduled routines arrived in March, while meeting recording, Chrome extension support and parallel helpers were part of the May update. Those are existing capabilities, not October 26 launches. Strawberry had already opened its beta to everyone in February.

So the practical question is less about whether a companion has a new face and more about whether the browser can finish a recurring piece of work without creating another review burden.

## How Strawberry's browser agent works

Strawberry's [browser-agent documentation](https://strawberrybrowser.com/browser-agent) describes a system that plans a task, reads pages, switches tabs and interacts with websites. It can also use a connected app when that is a better route than browser interaction. Logged-in work happens inside the browser session the user already has.

Consider a small vendor comparison. The useful output is a sheet containing prices, delivery dates and the sources behind them. A chat answer still leaves someone to move that information into a working document. Strawberry is designed to handle those intermediate steps as well as the research.

That does not establish that it can reliably complete every website workflow. The product page describes its intended behavior, not an independent success-rate evaluation. For a first trial, a bounded task with a result you can check is more informative than asking it to take over an entire business process.

## Strawberry browser pricing

Strawberry's [individual monthly plans](https://strawberrybrowser.com/pricing) currently look like this:

| Plan | Monthly price | Included monthly credits |
| --- | --- | --- |
| Free | $0 | 500 |
| Intern | $20 | 6,000 |
| Part-Time | $85 | 25,000 |
| Full-Time | $250 | 75,000 |

Ordinary browsing does not consume credits. Companion work, including research and drafting, does. The free plan includes Efficient and Default modes; paid individual plans add Expert mode and model selection.

Monthly plan credits do not roll over. Purchased top-ups cost $0.01 per credit, or $10 for 1,000, and do roll over. These are application credits, not a published per-token API rate.

**A credit allowance is not a task allowance.** The pricing page does not promise a fixed number of completed vendor comparisons, CRM updates or research reports. Before moving a workflow onto a paid plan, run a representative example and check both its credit consumption and the time needed to verify the result. A low subscription price can still be poor value if each output needs substantial repair.

## The automation limit worth checking first

The clearest operational restriction is in Strawberry's [routine documentation](https://strawberrybrowser.com/features/routine-automation): routines run while Strawberry is open on your computer. Work that becomes due while it is closed waits until you reopen it.

**Scheduled does not mean independently hosted.** A morning briefing can be a good fit if the browser is running. A time-sensitive process that must execute regardless of your laptop's state needs a different reliability assumption.

Routines can start on a schedule or an event in a connected app, and each run appears in a chat for inspection. The documentation also describes approval points before messages are sent or important records are changed. Those checkpoints are useful, but they do not remove the need to review the proposed action.

## Who should try Strawberry 1.0

Strawberry lists macOS and Windows availability. Its [current privacy policy](https://strawberrybrowser.com/legal/privacy-policy) says the service is for business use, not personal, family or household use. The strongest initial fit is a business user whose repeated work crosses several browser tabs: gathering information, comparing it and moving the result into another tool.

The free tier offers a way to test that specific handoff before committing a broader workflow. The question to watch is whether reusable instructions and retained context reduce supervision over repeated runs. That is a more useful measure of an AI companion than how convincing its first demonstration looks.

I’m Paras. I work at the intersection of AI, engineering and distribution.

I help technical founders turn product launches and engineering updates into clear stories that developers and founders want to read and share.

[Book a 30-minute call](https://cal.com/paras-madan/30min).