NBot
NBot builds AI trackers that continuously monitor news sites, newsletters, RSS feeds, niche blogs and social discussions on any topic you describe in plain language, returning cited summaries, a chat-driven feed and daily AI podcasts.
What is NBot?
NBot is an AI content curation platform built around what it calls AI Trackers: standing agents that watch the web for a topic described in ordinary language. The site presents the workflow in three steps. First, Describe — you tell NBot what interests you instead of assembling boolean strings or maintaining RSS subscriptions, and the assistant claims to understand niche technical terminology, identify leading sources on its own and suggest adjacent topics to widen coverage. Second, Discover — the tracker monitors news outlets, newsletters, RSS feeds, niche blogs and social media discussions, extracts core arguments, highlights new data points and attaches citations so every claim can be checked at source. Third, Refine — a tracker is not a static filter but an active agent, and a built-in chat lets you question the feed or redirect it mid-flight.
Two kinds of tracker exist: AI Trackers, which curate, and Subscription Trackers, which carry raw feeds. Sources attach as X accounts, YouTube channels and Reddit accounts, with per-tier limits; websites and RSS are declared when a tracker is created rather than through the settings dialog. Configuration reaches unusually far into editorial behaviour — curation strategy, tone rules, search terms, followed domains with optional geolocation, followed subreddits, and persona attributes including political standpoint, gender and base city that shape the perspective applied to content. Refresh frequency is a slider from two to twenty-four runs per day, with the credit cost displayed before you commit.
Output travels well beyond the web app. Each tracker can generate a daily AI podcast with a male or female narrator, push items to Telegram as individual messages or a single digest, and be exposed through a REST API and an MCP server offering fourteen tools across curator management, content and task monitoring, usable from Claude, Cursor or any MCP-compatible client. Public trackers can be followed or published, and the homepage carries a live community feed timestamped to the minute. Native iOS and Android applications complete the web interface. The publisher is Local AI, Inc., a Delaware company based in Mountain View, California.
What it does
- Describe a topic in plain language and let the AI build a tracker around it
- Monitor news outlets, newsletters, RSS feeds, niche blogs and social discussions continuously
- Read AI summaries that explain why an item matters, each linked to its original source
- Chat with a feed to ask questions or re-steer what the tracker collects
- Listen to a daily AI-generated podcast of the tracked content
- Receive matching items on Telegram, individually or as a single daily digest
- Plug trackers into Claude, Cursor or any MCP client through the API
When to use NBot / When not to
A quick filter to help you decide if NBot is the right fit.
When to use NBot
- Journalists and editors who need to follow a beat across hundreds of outlets without drowning in headlines
- Analysts and investors tracking specific themes such as monetary policy, crypto, precious metals or uranium
- Researchers and academics monitoring niche technical literature, including preprint servers and specialist journals
- Product, marketing and competitive intelligence teams watching a market, a category or a set of competitors
- Developers who want a curated content stream inside their own tools through the API or an MCP-compatible assistant
When not to use NBot
- Organisations bound by GDPR obligations: the regulation is never mentioned, no DPA is published and no EU representative is named
- Teams that need data residency outside the United States, since all processing is described as taking place there
- Users unwilling to have their identifiers and network activity sold or shared with advertising partners, as the privacy policy declares
- Buyers who require a published security certification, a subprocessor list or a service level commitment below the Enterprise tier
- People looking to add a website or a raw RSS feed as a subscription source, which the settings dialog explicitly does not offer
How to use NBot
A typical end-to-end flow, from setup to results.
- Open the site and click Login in the top navigation
- Sign in with a Google, Apple or X account
- Land on the homepage, which shows the For You feed, your trackers and the Explore section
- Click New Tracker in the navigation bar
- Describe the topic you want to follow in the chat interface
- Let the assistant propose search terms, sources and preferences
- Review the configuration card and confirm the settings
- Wait a few minutes for the tracker to return its first curated items
- Open the gear icon to adjust sources, podcast voice and daily run frequency
- Connect Telegram or generate an API key from Settings to receive the content elsewhere
Pros & Cons
Pros
- A tracker is set up by describing a topic, with no boolean queries or RSS plumbing to maintain
- Every summary carries its citation and a direct link, so a claim can be verified immediately
- The free tier is permanent, needs no credit card and is advertised with full feature access
- Pricing is public and legible: four tiers and a credit table stating what each action costs
- Nine documentation pages cover creation, configuration, feeds, podcasts, Telegram, credits and the API
- A standard MCP server makes the curated content usable from Claude, Cursor or any compatible client
- Refresh frequency is adjustable from two to twenty-four runs a day, with the credit cost shown before confirming
Cons
- No GDPR framework at all: the regulation is never mentioned and no EU representative is designated
- The privacy policy declares CCPA sales and shares of identifiers and network activity to advertising partners
- Browser Do Not Track signals are explicitly not honoured, and the advertising opt-out runs through a mobile app setting
- No data processing agreement, no subprocessor list and no published security certification
- Data is processed in the United States only, with no regional hosting option
- API access and the MCP server are locked to the Pro plan
- Running out of credits makes trackers skip their next run, and daily credits never roll over
Pricing & Plans
A permanent free plan is available and requires no credit card; it provides twenty daily credits, two AI Trackers and, according to the pricing page, full feature access. The lowest paid entry point is the Plus plan at USD 9.99 per month, reduced to USD 8.33 per month on annual billing (USD 99.99 per year). The Pro plan is USD 19.99 per month, or USD 16.67 per month billed annually (USD 199.99 per year). An Enterprise tier is quoted on request. Payments are processed by Stripe.
- Free — no charge — 20 daily credits
- 2 AI Trackers
- 30 Subscription Trackers
- 3 updates per day
- 3 search terms per tracker
- standard data refresh
- Plus — USD 9.99 per month
- or USD 8.33 per month billed yearly (USD 99.99 per year) — 20 daily credits plus 1
- 000 monthly credits
- 10 AI Trackers
- 100 Subscription Trackers
- 6 updates per day
- 10 search terms per tracker
- fresher data refresh
- Pro — USD 19.99 per month
- or USD 16.67 per month billed yearly (USD 199.99 per year) — 20 daily credits plus 2
- 500 monthly credits
- 30 AI Trackers
- 100 Subscription Trackers
- 8 updates per day
- 50 search terms per tracker
- freshest data refresh
- Enterprise — custom pricing — custom credit allocation
- unlimited AI Trackers
- top priority updates
- custom API integrations
- SSO enforcement
- SLA guarantee
- dedicated support and onboarding
Data, GDPR & hosting
A consolidated view of how NBot handles your data.
GDPR overview
There is no mention of the GDPR anywhere on this site. Searching the full text and the archived HTML of every page collected returns zero occurrences of GDPR, EEA, European Economic Area or Article 27. This is silence rather than a refusal: the company neither claims compliance nor denies it. The framework actually invoked is United States state privacy law, chiefly the California Consumer Privacy Act as amended by the CPRA, with named rights for residents of Colorado, Connecticut, Virginia, Oregon, Utah and Texas covering knowing, accessing, correcting and deleting personal information. Requests go to a single address, compliance@localaiapp.com, under the subject Privacy Law Request, and identity verification comes first. Data is stated to be processed in the United States. No EU representative and no data protection officer is designated. The privacy policy is dated 14 April 2025.
Who owns the data?
Local AI, Inc. operates the service and collects account details, device and usage signals, general location and any content a user posts. Users keep the ability to erase their personal data by deleting their account from the settings, or by writing to the compliance address if that fails. The point deserving most attention is publicity rather than ownership: the policy states that a username, the general location declared at registration, liked content and created content such as posts, comments or video are available to the public. A content provider a user follows can also see that user's account name, general location and profile picture. General location can be excluded from public comments in the settings.
Reuse rights
The published documents describe how Local AI, Inc. uses the information it collects rather than granting users any right to reuse the platform's content. Declared purposes cover operating and improving the service, personalising content and advertisements, marketing, internal research and development, quality control, security and legal compliance. Advertising partners receive device and usage data through embedded SDKs so they can bid on inventory and measure campaigns, and browsing activity may be matched across a user's devices. Aggregated or de-identified data may be used for any purpose and disclosed to third parties, advertisers included. Under the CCPA the company states that it has sold or shared identifiers and network activity information over the past twelve months; that disclosure can be refused through the About setting in the mobile application. Do Not Track signals sent by a browser are explicitly not honoured, although a legally recognised opt-out preference signal is.
Data retention & training
Hosting summary
The privacy policy states plainly that NBot is located in the United States and processes user data there. No alternative country or region is offered, and no residency option appears anywhere in the published material. The company reserves the right to subcontract processing to service providers and business partners in countries other than the user's country of residence, the United States included, without naming them: no subprocessor list is published. Security is described in general terms only — administrative, technical and physical measures — and the policy explicitly declines to guarantee it, noting that no method of transmitting or storing information is completely secure. There is no trust or security page, and no certification such as SOC 2 or ISO 27001 is displayed. As an observation rather than a company statement, a DNS lookup carried out on 11 August 2026 resolved the domain to an address in Amazon's network located in Boardman, Oregon, consistent with the declared United States jurisdiction.
Things to keep in mind
Risks and trade-offs to weigh before adopting NBot.
- Your identifiers and network activity may be sold or shared with advertising partners, and the opt-out lives in a mobile app setting rather than on the web
- Content you create, your username and the general location you declare at registration are public by default
- No GDPR framework, no EU representative, no data processing agreement and no subprocessor list; all processing takes place in the United States
- A tracker that never stops running rewards constant checking, and the very design meant to beat information overload can quietly manufacture it
- Reading AI summaries instead of sources is comfortable and slowly erodes the habit of reading the source at all; the citations are there to be used
- Persona settings including political standpoint shape the perspective applied to your feed, which can harden a filter bubble rather than broaden it
- The terms disclaim any warranty on the accuracy of generated responses, cap liability at the higher of what you paid or USD 100, and impose arbitration with a class action waiver
Setup & Integrations
Technical difficulty
Very low for ordinary use. There is nothing to install: you sign in with a Google, Apple or X account and create a first tracker by describing a subject in a chat, with no query syntax to learn. The documentation says the first results arrive within minutes, and mobile applications are available on both stores. Only two steps call for any technical comfort, and both are optional: generating an API key and pasting a short JSON block into an MCP client, which the Pro plan gates, and authorising the Telegram bot from Settings.
Deployment
Apps stores
Integrations
Behind NBot
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is an AI Tracker?
Is there a free plan?
How much does the cheapest paid plan cost?
How do credits work?
Is there an API?
Can I get the content outside the website?
Is NBot GDPR compliant?
Is my data used for advertising?
What is the minimum age?
Can I delete my data?
Should you pick NBot?
NBot is a working product rather than a promise. The homepage carries a live feed timestamped to the minute with outbound links to third-party sources, nine documentation pages describe the product in detail, pricing is public across four tiers, and applications are published on both mobile stores. The idea it defends — describe a subject in ordinary language and let an agent maintain the watch, citation by citation — is executed with unusual depth of configuration, down to tone rules, followed subreddits and a refresh frequency priced in credits before you confirm it. The free tier is permanent and asks for no card, which makes the tool genuinely testable.
The reservation concerns not the product but what surrounds it. On personal data the published documents point away from what a European reader will expect: the GDPR is never mentioned, no EU representative is designated, no data processing agreement or subprocessor list exists, no security certification is displayed, and processing is described as taking place in the United States alone. More pointedly, the privacy policy states that identifiers and network activity have been sold or shared with advertising partners over the past twelve months, that browser Do Not Track signals are not honoured, and that the opt-out is reached through a setting inside the mobile application. Content a user creates, along with the general location declared at registration, is public by default.
The result is a tool that suits individual curiosity, editorial work and market watching very well, and that an organisation carrying regulatory obligations should examine closely before adopting. Local AI, Inc. is a Delaware company operating from Mountain View, California; the domain was registered in April 2025, so the service is young and its published commitments remain thin.
- Choosing a selection results in a full page refresh.
- Opens in a new window.