✦ Chrome Extension Manifest V3 β€’ Free & 100% Private β€’ Version 1.3.0

Twitter (𝕏) AI Assistant: Auto Reply & Post Creator Cloud & Nano AI

One-click context-aware AI replies and viral post generation directly beneath tweets on 𝕏. Powered by official Google Gemini Cloud API (gemini-3.1-flash-lite), on-device Gemini Nano, & silent sessions β€” sub-second latency, zero monthly fees, and 100% private.

βœ“ Free Cloud API & Local Nano
βœ“ 4 Modular Engines + Fallback
βœ“ 0 External Data Tracking
βœ“ Open Source on GitHub
E
Elena Rostova
@elena_devs

Running local LLMs inside Chrome via Gemini Nano is a game changer. Sub-second response times, zero API bills, and complete privacy since data never leaves your device. Are you moving to local AI yet?

54
28
412
38.4K
Replying to @elena_devs
⚑ Gemini Nano · 180ms · Local GPU
Select Reply Tone: Click any tone to test live AI drafting β–Ύ
✦ AI Reply
β–Ύ

From Tweet to High-Signal Reply in Under 2 Seconds

No API dashboards, no credit cards, and no workflow interruption. Everything happens in-place right on your Twitter feed.

01

Choose Your Engine in 1 Click

Connect via Google Gemini Cloud API (free key from Google AI Studio), on-device Gemini Nano (0 tabs, offline), or silent Google session cookies.

Instant Setup Cloud API, Nano, or Web
02

Browse 𝕏 with Floating HUD

Scroll your home feed, notifications, or lists. The extension automatically detects reply composers and provides a floating on-screen dock.

In-DOM + HUD Draggable dock + Native buttons
03

AI Reply or Viral Post Creator

Choose from 6 contextual reply tones (Thoughtful, Witty, Support, Counter-Point) or 7 standalone viral tweet archetypes.

Tones & Styles ⚑ πŸ‘ πŸ’‘ πŸ˜‚ πŸ”₯ ❓ πŸš€
04

Review & Post with Guardrails

Generates clean, human copy without internal thinking preambles or asterisks. You retain full editorial control before posting.

Safe & Clean Preamble-free clean output

Four High-Performance AI Engines. You Choose.

Switch effortlessly between official cloud REST models, on-device local hardware, headless web sessions, or automated tab execution based on your workflow.

πŸ”„
Intelligent 4-Engine Fallback Pipeline Zero Downtime
If an engine encounters a rate-limit or temporary timeout, the extension automatically cascades to keep generation seamless.
☁️ Cloud API
→
⚑ Gemini Nano
→
🌐 Headless Web
→
πŸ—‚οΈ Gemini Web Tab
⚑

Gemini Nano

Chrome's built-in on-device AI. Executes directly on your PC's GPU or NPU without sending a single byte over the web.

  • Browser Tabs Opened: 0 tabs (Silent)
  • Generation Speed: < 250ms (Instant)
  • Privacy Level: 100% On-Device Offline
  • Cloud Dependency: None (No internet needed)
🌐

Headless Direct

Direct background requests utilizing your active Google login session cookies. Full Gemini cloud intelligence with zero clutter.

  • Browser Tabs Opened: 0 tabs (Headless)
  • Generation Speed: ~650ms
  • Privacy Level: Direct to Google (No 3rd Party)
  • API Key Needed: None (Free Session)
πŸ—‚οΈ

Web Tab Automation

Lightweight tab automation that interacts with gemini.google.com chat DOM. Features smart auto-close so tabs vanish immediately after pasting.

  • Browser Tabs Opened: 1 Temporary Tab
  • Auto-Close Tab: Yes (Vanish on completion)
  • Generation Speed: ~1.5s
  • Compatibility: Universal Chrome
Capability ☁️ Cloud API (Recommended) ⚑ Gemini Nano (Local) 🌐 Headless Direct πŸ—‚οΈ Web Tab Automation
API Key or Credit Card Free Key (Zero Billing/Card) Not Required Not Required Not Required
Browser Tabs Opened 0 Tabs (Direct REST) 0 Tabs 0 Tabs 1 Auto-closing tab
Latency < 220ms (Ultra-Fast) < 250ms (Zero lag) ~650ms ~1.5s
Multi-Model Selection Yes (Flash Lite, 2.5, 2.0, Pro) Nano Prompt API Standard Web Model Standard Web Model
Works Completely Offline No (Internet required) Yes (100% On-Device) No No
Third-Party Server Intermediary Zero (Direct to Google) Zero Zero Zero

Built for Serious Creators & Power Users

Every feature is designed to elevate your Twitter presence with authentic, high-signal conversations while eliminating repetitive drafting.

🎭

6 Distinct Tones of Voice

Switch effortlessly between Quick & Natural, Agree & Supportive, Thoughtful Insight, Witty Humorous, Counter-Point, and Question to match the conversation mood.

✍️

AI Post Creator & 7 Archetypes

Beyond replies: craft viral standalone tweets. Use Topic Prompt mode to start from scratch, Draft Polish mode to refine rough thoughts, or use the dedicated Popup Post Creator.

πŸ–₯️

On-Screen Floating HUD Widget

A draggable, collapsible dock sitting directly on your Twitter feed. Shows live status of all 4 engines, quick tone switcher, and an inline multiline custom prompt editor with quick-save.

πŸ”„

Intelligent 4-Engine Fallback

Never get blocked by rate limits. Automatically cascades across Cloud API → Gemini Nano → Headless Web → Gemini Web Tab to guarantee 100% reply uptime.

βš™οΈ

Custom Rules & Strict Guardrails

Define strict rules (e.g. "no emojis", under 180 chars, brand voice). Automatically strips thinking scratchpads, preambles, asterisks, and multi-option headings.

πŸ”’

100% Client-Side Privacy & Direct REST

No middleman servers, no analytics, no database. API keys and preferences are stored strictly in local browser storage (chrome.storage.local).

πŸ›‘οΈ

Your Feed and Ideas Stay Yours. Always.

Unlike other AI extensions that route your tweets through unknown third-party proxy servers, our extension is 100% client-side and completely transparent.

🚫

Zero Third-Party Servers

We do not operate servers, middleware, or relays. Requests run either on your local hardware (Gemini Nano) or directly between your browser and official Google endpoints.

πŸ•΅οΈ

Zero Telemetry & Tracking

No Google Analytics, no PostHog, no Mixpanel, no crash report beacons. We have zero interest in tracking who you reply to or what you write.

πŸ”‘

Local Secure Key Storage

Your Google AI Studio API key and account cookies are stored strictly in your local device sandbox (chrome.storage.local) and never transmitted to any third party.

πŸ“–

100% Open Source

Every line of code is open source and licensed under MIT. You can audit the repository, verify every permission, or run it directly from source unpacked.

Manifest V3 Permissions Disclosure:

storage Saves your preferred engine, selected Cloud API model, AI Studio key, reply tone, and custom prompt rules locally.
tabs & scripting Allows injecting AI Reply and Post buttons into Twitter (x.com) composers and managing the Gemini automated tab.
offscreen Accesses Chrome's built-in Prompt API (Gemini Nano) from the Service Worker context and handles clean clipboard interactions.
declarativeNetRequest Enables Headless Direct requests to communicate directly with your Google session without full webRequest interception.
host_permissions Strictly scoped to twitter.com, x.com, gemini.google.com, and generativelanguage.googleapis.com (official Google Gemini API).

See the Extension in Action

Minimal, unobtrusive, and designed to look right at home with Twitter’s modern dark theme.

Twitter AI Auto Reply Extension Feature Marquee Extension Popup Settings and Post Creator Interface Twitter AI Assistant On-Screen Floating HUD on x.com

Frequently Asked Questions

Everything you need to know about engines, pricing, permissions, and compatibility.

Do I need a paid API key or credit card to use this?

No, absolutely not. You have complete freedom: you can obtain a 100% free API key from Google AI Studio (no billing, no credit card required, with generous daily free tier limits) for ultra-fast Cloud API performance, run 100% offline with Gemini Nano (zero keys), or use the Headless Direct engine with your active Google login session cookies.

Which AI models can I choose from?

When using the Google Gemini Cloud API engine, you can select from: Gemini 3.1 Flash Lite (default recommended, ultra-fast latency with sub-second generation), Gemini 2.5 Flash, Gemini 2.0 Flash, Gemini 1.5 Flash, and Gemini 1.5 Pro (for deep analytical reasoning). You can switch models anytime in the popup settings.

What is the on-screen Floating HUD widget?

The Floating HUD is an optional draggable control dock that appears directly on your Twitter (𝕏) feed. It displays live status chips for all 4 engines, lets you quick-switch reply tones and post archetypes, offers an inline multiline editor to update custom prompt rules on the fly, and lets you draft standalone tweets. You can drag it anywhere or toggle it off in the extension settings.

How does the Intelligent Fallback cascade work?

If your active engine encounters a temporary network error, rate limit, or model download delay, the extension automatically cascades through available backup engines: Cloud API → Gemini Nano → Headless Web → Gemini Web Tab. This guarantees uninterrupted tweet and reply drafting.

Is this extension completely free?

Yes! The extension is 100% free and open source under the MIT License. There are no paywalls, token limits, or premium subscriptions. If you enjoy the tool, you can support its development through Trakteer ID or Buy Me a Coffee.

How do I check if Gemini Nano works on my Chrome?

You can run a quick 5-second check: press F12 in Chrome, switch to the Console tab, and paste 'LanguageModel' in window || 'ai' in window. If it returns true, your browser supports built-in AI! If not, you can enable chrome://flags/#prompt-api-for-gemini-nano, or simply use the Cloud API or Headless Direct engines which work immediately on every setup.

Will my Twitter (𝕏) account get banned or flagged?

No. The extension does not post automated bot spam. It acts as an editorial copilot: it drafts an intelligent response and places it in your composer for your review. You choose when to edit, polish, and hit the Tweet button yourself.

Can I customize the reply language and length limits?

Yes! Both in the extension popup and on the on-screen Floating HUD, you can specify custom prompt instructions. You can instruct Gemini to always answer in a specific language (e.g. English, Indonesian, Spanish, French, Japanese), enforce "no emojis", keep replies under 180 characters, or instruct it to adopt a specific brand voice.

Start Writing High-Signal Replies on 𝕏 in Seconds

Join developers, creators, and power users who generate engaging replies effortlessly. Ultra-fast Cloud API, on-device Nano, and 100% client-side privacy.