ChatterKB
ChatterKB is a team workspace that turns uploaded files and connected tools into a searchable AI knowledge base, then runs plain-English workflows and refreshable dashboards on top of it. A free plan precedes paid tiers from USD 20 monthly.
What is ChatterKB?
ChatterKB is an AI workspace built around three connected pieces that the homepage sums up as Think, Do, Show. The first is a knowledge base. You upload PDF, Word, Excel, PowerPoint and text files, or connect existing sources: Google Drive and Notion through OAuth, Slack channels in one click, and YouTube videos or playlists by pasting a URL with no authentication at all. Each file produces pre-built insights you can browse, and the content is stored as editable markdown rather than as a frozen attachment. Synchronisation is deliberately manual, so your originals stay untouched while ChatterKB works on a copy.
The second piece is chat. You ask questions in ordinary English and the assistant answers from your knowledge base and integrations, with shared memory so context carries across teammates and threads. A Save button promotes any useful answer into the knowledge base. Crucially, the model is not fixed: Claude, ChatGPT, Google Gemini, Perplexity, Grok, Meta, DeepSeek and Amazon Nova can be swapped task by task, and the enterprise page claims more than twenty models in total.
The third piece is automation. You describe a recurring task in plain English, ChatterKB generates the steps, and you then tune each one — model, tools, memory, conditional branches — without scripting. Workflows run on a click or on a schedule, and pause for human input when something goes wrong. Their outputs can be pinned to a Board, refreshed at any time, and shared with colleagues or clients. A public library of twelve ready-made workflows covers SEO keyword discovery, ad copy evaluation, market intelligence, company research, YouTube analysis and client onboarding.
Model Context Protocol support connects further tools: Zapier, Notion, HubSpot, Google Analytics, monday.com, Airtable, Google Workspace and Reddit. For larger buyers there is a bespoke retrieval-augmented offer with cited answers, granular access controls and deployment in your own AWS, Azure or GCP account or on-premises. The publisher is Beatspoken LLC. There is no public API and no mobile application.
What it does
- Answer questions in plain English from your own files, notes and connected tools, grounded in that content rather than invented
- Generate a multi-step workflow from a written description, with no code and no flowchart, then run it on demand or on a schedule
- Build a knowledge base from PDF, Word, Excel, PowerPoint and text uploads, plus Google Drive, Slack, Notion and YouTube transcripts
- Pin reports, summaries and charts to a Board and press refresh to recompute them against the latest content
- Switch model per task across Claude, ChatGPT, Gemini, Perplexity, Grok, Meta, DeepSeek and Amazon Nova
- Save any useful answer into the knowledge base as editable markdown that the whole team can reuse
- Publish a branded public or private knowledge base for clients, with your own logo, colours and custom CSS
When to use ChatterKB / When not to
A quick filter to help you decide if ChatterKB is the right fit.
When to use ChatterKB
- Marketing and creative agencies that keep client knowledge scattered across drives, inboxes and personal notes, and lose it when someone leaves
- Small, non-technical operations teams that want recurring reporting and research automated without learning a node editor
- Independent consultants and solo operators, who fit the single-seat Free, Plus, Pro and Pro XL tiers exactly
- Regulated organisations in life sciences, finance, legal or the public sector that need retrieval with citations inside their own AWS, Azure, GCP or on-premises environment
- Teams that refuse to commit to one model vendor and want to switch between Claude, ChatGPT, Gemini, Perplexity and Grok task by task
When not to use ChatterKB
- Developers looking for a public API: the site publishes no API documentation, and both /api and /docs return a hard 404
- Anyone who needs a mobile or offline client, since ChatterKB ships only as a browser application with no iOS, Android or desktop build
- Genuinely collaborative teams on a modest budget, because every plan below Business is capped at a single user and Business starts at USD 250 per month by quote
- Organisations that require an independent security attestation before signing, as no SOC 2, ISO 27001 or HIPAA certification is published anywhere on the site
- Teams that need connected sources to update themselves: synchronisation is manual by design, so nothing refreshes until someone triggers it
How to use ChatterKB
A typical end-to-end flow, from setup to results.
- Create an account at app.chatterkb.com; the free plan needs no credit card
- Create or open a knowledge base, which is the container everything else hangs from
- Upload your documents — PDF, Word, Excel, PowerPoint or plain text — in batches if needed
- Connect existing sources: authorise Google Drive and pick files, link Notion by OAuth, import Slack channels, or paste a YouTube URL
- Add Model Context Protocol servers if you need Zapier, HubSpot, monday.com, Airtable or Google Analytics data in the mix
- Ask your questions in plain English in the chat, choosing the model that suits the task
- Press Save on any answer worth keeping, then refine it as markdown inside the knowledge base
- Describe a recurring task in plain English and let ChatterKB draft the workflow steps
- Tune each step — model, tools, memory, conditional logic — then run it once or schedule it
- Pin the results to a Board, refresh whenever you need current figures, and share it with your team or your client
Pros & Cons
Pros
- No model lock-in: you choose per task among more than twenty models rather than inheriting one vendor's roadmap
- Workflows are written in plain English, which genuinely opens automation to people who will never touch a node editor
- An unambiguous, written commitment that customer data is never used to train, fine-tune or improve models, repeated in the DPA
- Compliance documentation that is complete and public: a Trust and GDPR page, an Article 28 DPA, and nine named subprocessors with purpose, location and safeguards
- A permanent free plan with no credit card, useful enough to judge the product
- Full pricing transparency in static HTML, down to the credit cost of each individual model
- Real sovereignty options: deployment in your own AWS, Azure or GCP account or on-premises, with EU storage available on request
Cons
- Every plan below Business is limited to one user, despite the product being marketed around shared team memory; multi-seat starts at USD 250 per month by quote
- No public API and no client beyond the browser: the 94-URL sitemap contains no API documentation, /api and /docs both return a hard 404, and there is no mobile, desktop or extension build
- Source synchronisation is manual by design, so a changed document does not reach the knowledge base until someone syncs it
- Credit consumption is hard to forecast, as a workflow costs roughly twelve times a chat message and usage grows with integrations, memory and complexity
- No published security certification — no SOC 2, no ISO 27001, no HIPAA — despite explicitly targeting healthcare, pharma, legal, banking and public sector buyers
- Every named subprocessor processes data in the United States, and EU storage is only a request that may be honoured where available
- The publisher is thinly documented: no postal address, no named leadership, no social presence, and a privacy policy that stops mid-way through the rights section
Pricing & Plans
A permanent free plan is available at no cost and without a credit card, offering one knowledge base and roughly 20 credits per month. The cheapest paid tier is Plus at USD 20 per month on monthly billing. Annual billing carries a 20% discount, and payments are processed through Stripe.
- 1 knowledge base
- 20 credits per month
- 1 user
- limited workflows
- Claude and Meta models with limited ChatGPT access
- and access to shared knowledge bases
- 3 knowledge bases
- 300 credits per month
- 1 user
- full workflows
- all models including Gemini
- Perplexity and Grok
- public and private knowledge base sharing
- Boards and Board sharing
- 5 knowledge bases
- 800 credits per month
- 1 user
- otherwise identical to Plus
- 10 knowledge bases
- 1
- 600 credits per month
- 1 user
- otherwise identical to Plus
- custom knowledge bases
- credits and user count
- and optionally unlimited storage
- unlimited chats
- dedicated support
- advanced analytics
- account management and enhanced security
- 20% discount on every paid tier
- annual plans are non-refundable after the first 30 days
- a simple chat costs 1 to 5 credits and a workflow around twelve times a chat message
- published per-model costs range from 0.10 credit for GPT-5 mini
- Grok 3 Mini and Amazon Nova Lite up to 1.90 credits for Claude Opus 4.6
- Visa
- Mastercard and American Express cards
- PayPal
- and ACH for annual subscriptions
Data, GDPR & hosting
A consolidated view of how ChatterKB handles your data.
GDPR overview
GDPR treatment is unusually well documented for a vendor this young. A dedicated Trust, Security and GDPR page, last updated 30 December 2025, states compliance explicitly and names Articles 28 and 32. ChatterKB positions itself as processor and the customer as controller. A data processing agreement is published at /dpa under Article 28 of Regulation (EU) 2016/679 and incorporates the European Commission's Standard Contractual Clauses by reference. A named subprocessor list is public, with purpose, data location and safeguards for each of the nine entries. Data subject rights, breach notification without undue delay, and EU-based storage on request where available are all addressed, and privacy@chatterkb.com is given as the contact. Two gaps remain: no Article 27 EU representative is named, and no independent certification is published.
Who owns the data?
You keep ownership of everything you upload. Clause 3 of the terms of service, last updated 22 April 2026, states that you retain ownership of your files and content, while granting Beatspoken LLC a non-exclusive, transferable, sub-licensable, royalty-free and worldwide licence to use, display, store and process that content, but only in order to provide and improve the service. The platform itself, its features and its original content remain the property of Beatspoken LLC. Under the data processing agreement, ChatterKB acts as processor and you as controller, and clause 6 forbids it from using personal data for model training, fine-tuning or product improvement outside your own account.
Reuse rights
Because ownership stays with you, nothing in the terms requires you to ask permission before reusing your own material or the outputs the assistant produces from it. Answers can be opened, edited and saved back into the knowledge base, which is stored as editable markdown that the site itself compares to a Notion page, and the trust page confirms that you may request a full export of your data at any time. Original files stay untouched in Notion, Google Drive or Slack, since ChatterKB processes a copy. From the Plus tier upward you can share knowledge bases publicly or privately and publish branded knowledge bases carrying your own logo, colours and CSS to clients. The limits are conduct-based rather than ownership-based: clause 4 of the terms prohibits unlawful or harmful content, infringement of third-party intellectual property, and anything that disrupts the service or its networks.
Data retention & training
Hosting summary
ChatterKB runs on cloud infrastructure and operates no datacentre of its own. The published subprocessor list, dated 30 December 2025, names Google Cloud Platform for application hosting, databases and networking, Amazon S3 for object storage of customer documents and retrieval files, and Amazon Cognito for authentication — all three located in the United States. Model inference runs on Amazon Bedrock, OpenAI, Google Vertex AI, Perplexity AI and xAI, again all in the United States, under Standard Contractual Clauses with transient processing and no model training. Encryption is applied in transit over HTTPS/TLS and at rest for databases and object storage, with logical tenant isolation and a separate schema per knowledge base. Passwords are not handled directly, as credentials sit with the identity provider. EU-based storage for application data and source documents may be requested where available, but it is not the default. Enterprise customers can instead deploy in their own AWS, Azure or GCP account or on-premises, which the vendor recommends for regulated industries. Note that the marketing site itself is served from a GitHub Pages edge node, which says nothing about where customer data lives.
Things to keep in mind
Risks and trade-offs to weigh before adopting ChatterKB.
- The content licence is broad: non-exclusive, transferable, sub-licensable and worldwide, granted to provide and improve the service. Read clause 3 before uploading commercially sensitive or client-confidential documents
- All named subprocessors process data in the United States, including five model providers. Transfers out of the EEA are covered by Standard Contractual Clauses, but they are still transfers out of the EEA
- No independent security attestation is published, so every security claim rests on the vendor's own statement rather than on an audit
- Clause 6 of the terms allows the publisher to terminate or suspend your access immediately, without prior notice or liability, for any reason
- Retention is never quantified: data is deleted within a defined retention period, with no number of days given anywhere
- Dependency builds quietly. Once the knowledge base becomes the team's reference, the only stated exit is an export on request, with no portability commitment attached
- Cited, sourced answers still come from a language model. Grounded retrieval narrows the room for invention, it does not remove the need to check what matters
Setup & Integrations
Technical difficulty
Day-to-day setup is easy. Signup needs no credit card, everything happens in the browser, and workflows are described in plain English — the vendor states the product is designed specifically for non-technical users. Connecting sources is equally light: OAuth for Google Drive and Notion, one click for Slack, nothing at all for YouTube. Two things are harder. Model Context Protocol servers require configuration, and local servers need the local MCP bridge. Enterprise deployment into your own AWS, Azure or GCP account or on-premises is an infrastructure project, though the vendor assists. Manual synchronisation demands discipline rather than skill.
Deployment
Integrations
Supported languages
Behind ChatterKB
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement ChatterKB.
Frequently asked questions
Is there a free plan, and what does it include?
How many people can use one subscription?
How does pricing actually work?
Which AI models can I use?
Is my data used to train the models?
Where is my data hosted?
Do the connected sources stay up to date automatically?
Is there an API or a mobile app?
Can I run it on my own infrastructure?
What are the refund terms and the minimum age?
Should you pick ChatterKB?
ChatterKB is a coherent product rather than a chat window with features bolted on. Knowledge feeds workflows, workflows feed dashboards, and the refusal to lock you into one model vendor is a real argument rather than a slogan. What stands out most, though, is the paperwork: a young publisher has produced a Trust and GDPR page naming Articles 28 and 32, an Article 28 data processing agreement with Standard Contractual Clauses, and a subprocessor table listing all nine third parties with purpose, location and safeguards. The written commitment that customer data is never used to train or fine-tune models is unambiguous and repeated contractually. Pricing transparency goes further than most, down to the credit cost of each individual model.
Two reservations deserve weight. First, the seat count: a product sold on shared team memory caps Free, Plus, Pro and Pro XL at a single user, so genuine collaboration means the Business plan at USD 250 per month by quote. Anyone budgeting from the public grid should check this before anything else. Second, the vendor is thinly documented — no postal address, no named leadership, no social presence, no independent security certification — while explicitly courting healthcare, pharma, legal, banking and public sector buyers who normally demand exactly that. The domain was registered in October 2024 and first archived a month later, so this is an early product, and the short privacy policy that stops mid-way through the rights section shows it.
For an agency, a consultant or a small operations team willing to work within one seat and manual syncing, the value is easy to see and the free plan costs nothing to test. For a regulated enterprise, the compliance documents are a good start but not a substitute for the audit evidence that is still missing.
- Choosing a selection results in a full page refresh.
- Opens in a new window.