Product link: https://trustmrr.com/startup/aera-browser?ref=eric-kwok-b654e8 (with referral tracking — also at the end)

Aera Browser Deep Dive: The Browser That Works for You — How $20/Month Sells Your Busywork

1. Introduction: The Browser Is the Last Moat

The AI automation space is crowded: ChatGPT Scheduled Tasks, Claude Computer Use, BrowserBase headless clouds, Puppeteer script farms. Aera Browser, launched Dec 2025, picked the dumbest and smartest path — no headless cloud, no plugin, just a Chromium browser that works on a schedule.

One-liner: Aera is a browser that executes. Describe one recurring job in plain English and Aera runs it on a schedule in your own signed-in browser: reading pages, pulling numbers, filling forms, reporting what changed. Wake up to finished work.

As of 2026-08-26 TrustMRR snapshot: $343 MRR, 9 paid subscriptions, ~1,700 users, $3,635 lifetime, Domain Rating 9. Small numbers, pure sample: one US indie hacker, Stripe-verified, Chromium base, MCP, local-first. A textbook early micro-SaaS monetization slice.

2. Project Overview

Aera = Chromium real browser + plain-English scheduler + MCP connector + local-first storage. Homepage tagline: The browser that does the work. TrustMRR tags: AI / Productivity / Utilities. Audience: Developers, AI enthusiasts, workflow power-users.

Core Capabilities

Where It's Dependable (and Not)

• Dependable: reading, watching for changes, pulling numbers, triaging inbox, filling ordinary forms on a schedule.
• Not dependable: rich-text / code editors — the agent damaged content in testing; don't use for drafting docs or editing code.
• One scoped routine > long cross-site chain.
• No autonomous checkout — intentionally no payments.

3. Design Philosophy: 5 Principles

1. Local-first, but Honest About Inference

Your data stays local. Inference does not — unless you self-host. History, chats, tasks are local SQLite; paid inference goes via OpenRouter to Anthropic/OpenAI/Google; Free goes to your Ollama. The security page lists every exfiltration and admits no SOC 2 / no ISO 27001 — engineering honesty as GTM.

2. Your Browser, Not a Bot Farm

Drives your real Chromium profile with your signed-in sessions. No headless fingerprint, lower ban risk, zero migration cost.

3. One Sentence > One Selector Set

Re-reads page semantics each run instead of replaying selectors — resilient to redesigns. Trade-off: non-deterministic; the site says "we won't tell you it never fails."

4. Scheduling Is First-Class

Describe → Schedule → History → Notify. Fail-fast with step-by-step logs you read instead of stack traces.

5. Open Standards, No Lock-in

Chromium + MCP + OpenAI-compatible endpoints. Free forces self-hosting to push upgrades; paid sells hosted frontier models.

4. Detailed Tutorial: 7 Steps to Your First Auto-Worker

Step 0 — Prerequisites

Any OS that runs Chrome (4GB min, 8GB for heavy use). Free: install Ollama first. All plans need an Aera account.

Step 1 — Download & Import

Download from getaera.app/download → Import from Chrome → sign in (device fingerprint for referral tracking).

Step 2 — Configure Model

• Free: Settings → Models → http://localhost:11434
• Pro/Ultra: pick hosted models (GPT-4o, Claude 3.5, Gemini)

Step 3 — First Task in 60 Seconds

Sidebar: "Every day at 9am pull Stripe's yesterday revenue, MRR, new subs into Google Sheet row 1" → Run → Make recurring.

Step 4 — Schedule

Pick cadence, notifications, retries. Check Run History.

Step 5 — Connect MCP

Enable MCP Server → add to Cursor / Claude Desktop config → trigger browser automation from your IDE.

Step 6 — Skills Marketplace

Install community Skills or publish your own (counter increments; no browsing data uploaded).

Step 7 — Operate

Fail-fast, read logs, use local model for sensitive pages, expect updates every few days.

5. How Aera Makes Money

5 Replicable Plays: Free self-host as upgrade filter, $20 + $200 tiering, sell usage not seats (11x meter), subscription sells model not browser, MCP + Skills marketplace future take-rate.

6. Core User Analysis: How Much $ per User per Month?

Math: $343/9 = ARPU $38. 8 Pro + 1 Ultra = ~$360 ≈ actual (with annual discounts). All-Pro would be only $180 → at least one whale.

LTV (24mo): Pro $480 / Ultra $4,800. Conversion 0.5% is the bottleneck — 2% would be $1,360 MRR (4x).

7. Insights: 7 Takeaways

1. Browser is the last moat — real logged-in Chromium is the only stable wall-passer.
2. Sell the model, not the browser — container free, intelligence paid (same as Cursor).
3. Honesty is GTM — list risks before pricing.
4. 0.5% conversion is both opportunity and alarm — narrow Free Vision to push conversion.
5. MCP is growth leverage — embed in workflow, not ads.
6. Ultra $200 is a filter — price screens customers.
7. $15M ask is emotional, not valuation — 3644x ARR, a not-for-sale signal.

8. 6 Replicable Lessons

1. Browser free, model paid 2. Free must self-host 3. $20 anchor + $200 whale 4. First task succeeds in 60s 5. Logs are the product 6. MCP before ads

9. Risks

Stalled conversion, model cost swings, non-determinism, no compliance certs, solo-founder risk, big-tech squeeze.

10. Conclusion

Aera is an early but clean sample: 1 Chromium + 1 scheduler + 3 tiers + $343 MRR. Copy: container free, intelligence paid; Free self-hosted; auditable logs; scheduling first-class.

Product link: https://trustmrr.com/startup/aera-browser?ref=eric-kwok-b654e8

Sources: TrustMRR (Stripe-verified) + getaera.app (Pricing/Features/Use-Cases/Security/FAQ). Estimates marked.

Snapshot 2026-08-26. Analysis is author's view, not investment advice.