You're an advisor or investment professional who relies on FactSet for institutional-grade market intelligence. Now you want your AI assistant to reason about your actual client book instead of generic market commentary. To do that, it needs to read the accounts you've linked and shared through your own Truthifi login. That means giving your AI assistant your connected client accounts across custodians, plus the held-away positions clients hold outside what you manage, in one read-only place it can reason about, right next to the FactSet research you already run. Get started with Truthifi

For advisors on FactSet, market research is rarely the gap. FactSet's earnings models, factor data, and portfolio analytics give advisors and managers institutional-grade insight into markets and individual securities. What lives elsewhere is the client's complete financial picture: the 401(k) at a former employer, the outside brokerage, and the bank accounts and liabilities that factor into a household's real concentration. FactSet's portfolio analytics read the holdings someone loads into them, not a client's outside custodian logins, so that picture usually ends up scattered across statements and portals.1 FactSet organizes the world by security: a ticker, an identifier, a factor exposure. A client's question is organized by household: who owns what, in which account, at what cost. Answering it requires each custodian's position record, and FactSet doesn't hold that record for held-away accounts. Connect those accounts to Truthifi and you get a whole-book view that answers client questions in plain language, informed by everything FactSet helped you understand about the market.

1. Create a Truthifi account. Sign up at truthifi.com.

2. Link the accounts. Connect your clients' accounts across custodians plus held-away accounts they hold elsewhere via Truthifi's secure aggregator connection, where supported. By design, credentials don't pass through Truthifi.

3. Turn on MCP access. Live, read-only AI access via MCP is included on every current Truthifi plan, including the free tier. See truthifi.com/pricing for current Truthifi pricing.

4. Add Truthifi to your assistant. In your AI assistant's connector settings, add a custom connector pointed at https://api.truthifi.com/mcp, then authorize.

5. Turn it on and ask. Toggle Truthifi on and ask: Across my whole book, where is concentration or held-away exposure highest right now?

What is an MCP?

An MCP, short for Model Context Protocol, is the connection standard that lets your AI assistant read live data instead of you copying it over by hand. Anthropic introduced it as an open protocol, and support has since spread to other AI assistants; check your assistant's own documentation for current MCP support. In short, a "FactSet MCP" would mean a connector that lets an AI assistant query FactSet's own financial data directly from FactSet, separate from the bank and brokerage accounts Truthifi reads read-only. For how Truthifi's own server works, see Truthifi MCP.

Does FactSet have an MCP server?

Short answer: not officially, as of this October 2026 review. FactSet's published route into other systems is its developer API platform, and the MCP-style examples you'll find online are user-owned repositories rather than FactSet products.

FactSet hasn't shipped an official MCP server for conversational AI access as of this review (2026).1 Its developer API platform at developer.factset.com is built for engineering teams. It offers REST APIs, SDKs and the Open:FactSet Marketplace for data delivery, and it needs API credentials and a developer to wire it up. Those tools connect FactSet data to other systems, not to your AI assistant for natural-language questions about your book.

There are community-built examples. luseloso/Factset_mcp_agent_example shows an approach for deploying an agent using FactSet API connections in an MCP-style pattern, but these are user-owned repositories, not official FactSet products, and they're not designed for no-code advisor use.1

Because that gap persists, many advisors look instead for a no-code way to let their own AI assistant read the whole book, including the held-away accounts that sit outside any single platform, and answer questions out loud. That's a separate job from market research and portfolio analytics, and it's where Truthifi fits in.

About FactSet

FactSet is an institutional financial data and analytics company founded in 1978 by Howard Wille and Charles Snyder, and headquartered in Norwalk, Connecticut.1 It is an independent public company (ticker: FDS). For this page, the firmographics matter less than how each FactSet product line maps to the questions an advisor brings to an AI assistant:

• Estimates, fundamentals, ownership and M&A data: answers "what is this security worth, and who else holds it?" It doesn't answer "how much of it does my client hold across three custodians?"

• Portfolio analytics and multi-factor risk models: strongest on the portfolios you load into them. A held-away 401(k) shows up only if someone exports and maps it in.

• Economic and macroeconomic data sets: context for the investment thesis rather than household-level inputs.

• The Open:FactSet Marketplace and developer APIs: move FactSet data into other systems in your tech stack, which still takes a developer.

• AI-powered features within FactSet Workstation: natural-language search and generative AI document summarization, according to FactSet's own platform description. They point at FactSet's content, not at your clients' outside accounts.

According to FactSet's investor relations site, its client base includes investment managers, banks, hedge funds, and advisors.1 Client counts and coverage change, so confirm current capabilities and scale at factset.com. This page doesn't quote FactSet pricing; confirm subscription and API licensing costs directly with FactSet.

What FactSet leaves out is a plain-language view of what the client actually owns and owes across the custodians and accounts you connect: a household balance sheet your AI assistant can work from once those accounts are linked. Advisors who already lean on FactSet for research are well placed to gain from pairing it with a read-only, whole-book layer like Truthifi.

FactSet API: getting your FactSet data for AI

Does FactSet have an API? Yes. According to the FactSet developer portal, FactSet offers a formal developer API platform (developer.factset.com) with REST APIs, SDKs (Python, Java, .NET, etc.), and the Open:FactSet Marketplace for data delivery, but it is a developer API. Truthifi gives your AI assistant read-only access over MCP (the Model Context Protocol) to the client account data that sits beside your FactSet research. There's no code to write and no official FactSet API required. From one view, your assistant can pull up balances, holdings, and transactions across your connected accounts. Connect your accounts to your AI →

How Truthifi can help

Truthifi pulls your book into one read-only view your AI assistant can reason about: client accounts across 18,000+ institutions, decoded into clean, normalized data rather than raw feeds, from an independent source with no commissions, no ads, no advice. Truthifi corrects and reconciles what it reads, then runs 130+ automated diagnostics on each client's holdings (a company-reported count), so your AI reads figures that have passed Truthifi's own internal checks rather than any independent verification. According to Truthifi's own platform data, it reports 99.7% accuracy and monitors $6B+ in assets; both are company-reported figures, not an independent audit (see the Sources note for where they come from). Access at your bank and broker is read-only by architecture (see the safeguards below); the few Truthifi tools that change data do so only inside Truthifi, including refreshing your data on demand (3 credits per refresh). With your client accounts connected alongside your FactSet research, here's what you can ask:

• "Looking across my whole book, which clients carry the heaviest concentration in one stock or sector once held-away accounts are counted?"

• "Show held-away assets my clients hold outside my primary custodians that I'm not seeing in my main view."

• "Which clients are paying the most in total fees across my connected accounts, not just what I manage?"

• "Flag allocation drift over 5% from target across my connected households this quarter."

Here's what that looks like with a real client question you'll actually get asked. A client asks you, "given everything I own, am I too concentrated in tech?" FactSet can tell you how tech-heavy each fund in the managed account is. It can't see the 401(k) at a former employer, the rollover IRA opened elsewhere, or the taxable account at another custodian unless someone maps them in. Once those held-away accounts are connected, you can put the question to your AI assistant directly. Ask "where is this client actually concentrated when you count everything?" and the answer can shift. A household that looks diversified on the managed side can carry added tech exposure through an outside account that never rolled up into one view.

Truthifi handles read-only data: net worth, holdings, fees, concentration, and cross-custodian or held-away exposure. Market research, portfolio risk modeling, and investment analytics stay in FactSet. Truthifi adds the conversational, whole-picture client context layer your AI assistant reads alongside it.

Two entry points. Let Truthifi produce a decoded, cross-custodian read on your book, or connect the AI assistant you already use for live, conversational questions across your connected client accounts, read-only (see truthifi.com/pricing).

Is it safe to connect my accounts to Truthifi?

No online connection is ever completely risk-free, but these safeguards are built into how the connection works rather than promised (more at Truthifi security):

• Read-only by architecture. At your bank and broker, Truthifi's access is read-only. Your AI assistant can't move money, place trades or change anything there. A few Truthifi tools do change data inside Truthifi: adding or deleting assets you track by hand, and refreshing your data on demand. That limit is enforced by how the connection is built, not by a clause in some terms of service.

• Credentials don't pass through Truthifi. By design, logins are encrypted in the browser and passed straight to the aggregator, which returns a read-only token.

• Every request logged. Each tool call is counted, so you can review per-tool daily and monthly usage yourself.

• Bank-grade encryption in transit and at rest, with access controlled per account and revocable anytime (Connections → Agents → Revoke). Truthifi sits on top of, not in place of, your existing compliance, recordkeeping, and custody systems.

How to connect your book to your AI assistant

Setup runs entirely in your browser, with no command line and no developer, on MCP-capable assistants.

• Start by setting up a Truthifi account at truthifi.com.

• Link the accounts you want your AI assistant to see, your clients' accounts across custodians plus held-away accounts they hold elsewhere, through Truthifi's secure aggregator connection, where supported.

• Next, turn on MCP access: live, read-only AI access is included on every current Truthifi plan; see truthifi.com/pricing for current pricing.

• Add Truthifi to your assistant. The connector URL doesn't change from one assistant to the next: https://api.truthifi.com/mcp. In your assistant, open its connector settings (usually Settings → Connectors), add a custom connector named Truthifi, paste the URL, and authorize. Assistant-specific steps are in the connect guides.

• Turn it on and ask: "Across my whole book, where is concentration or held-away exposure highest right now?"

What you can ask your AI

• "Looking at every household, which clients hold the most concentrated position in one stock or sector once held-away accounts are factored in?"

• "Total fees by client across all connected accounts this year, not just what I manage, so I can flag the biggest for a fee review."

• "Which households are now more than 5% off their target allocation?"

• "By client, show me idle cash balances above $50k across my connected custodians."

• "For this client, compare what I manage against their held-away accounts and flag any overlapping concentration."

• "Where across my book is there overlapping single-stock exposure I might not be seeing in my primary view?"

Your book today vs. with Truthifi

Common questions about FactSet and AI

Does FactSet have an MCP or AI integration for my assistant? FactSet has not publicly released an official MCP server as of the reviewed date. FactSet does offer a developer API platform and the Open:FactSet Marketplace, tools built for engineering teams to connect FactSet data to other systems. To let your own AI assistant read your whole client book conversationally, including held-away accounts outside any single platform, a no-code connector like Truthifi provides that read-only layer.

Does Truthifi replace FactSet? No. FactSet delivers institutional market intelligence: earnings models, factor data, and portfolio risk analytics that inform your investment thesis. Truthifi is the additive, read-only context layer your AI assistant uses across the custodians and held-away client accounts you connect. They do different jobs.

Can the AI place trades or change anything? No. At your bank and broker, Truthifi's access is read-only by design, so your AI can't trade, move money, bill anyone or alter a setting there. The only changes it can make are inside Truthifi, such as adding or deleting assets you track by hand or refreshing data on demand.

What about held-away assets? That's the core reason advisors connect Truthifi: accounts a client holds outside your custody are read directly through the aggregator connection (where supported), so your AI can factor them into concentration, fees, and allocation. FactSet's research doesn't supply that picture on its own unless someone exports and maps those accounts into it.

Do I need to write code to set up Truthifi? No. Setup happens entirely in your browser, with no command line, API keys, or config files needed.

Which AI assistants does Truthifi work with? MCP-capable assistants that accept a custom connector can work with Truthifi. The connector URL doesn't change from one assistant to the next: https://api.truthifi.com/mcp. Step-by-step setup for specific assistants can be found in the connect guides linked below.

Truthifi and the data layer

FactSet's developer APIs and the Open:FactSet Marketplace move licensed market, fundamentals, and reference data into the other systems in a firm's tech stack. What a data layer like FactSet hands off, though, is data about securities, not the household account data Truthifi produces from your clients' institutions. Truthifi is the complementary layer that makes it Decoded, Verified, and yours: positions named and fees identified across your connected accounts, independently checked, and refreshed automatically each day. Once the household view is clean, your FactSet research can become more useful, because a factor or earnings-revision screen can be read against what each client holds in total rather than only what you manage.

Keep going

• See every connector: the financial MCP servers hub.

• How to ask your AI about your real portfolio data

• How to spot hidden investment fees

The short version

Held-away data rarely changes what you know about a security. It changes which households keep you up at night. Pick the household you consider most diversified and ask your assistant for its top three exposures with held-away accounts included. If that answer differs from your managed-side view, the gap itself is the agenda for that client's next review. Run the same question across your whole client book and the output stops being research and becomes triage: a ranked call list of the households whose real risk sits outside what you manage, something FactSet's research alone doesn't hand you. Connect your book to your AI →

See every connector on the financial MCP servers hub, or browse all connect guides.

Read next from the Truthifi blog

• Where is my book concentrated?

• What are clients actually paying in fees?

About the author

Truthifi Bot drafts these provider guides from Truthifi's data and public sources. Every page is reviewed by the Truthifi Team before publishing.

Sources

• "About FactSet", platform description, API/marketplace, community repos, and firmographics sourced from FactSet's own site (factset.com), FactSet investor relations (investor.factset.com), and the FactSet developer portal (developer.factset.com); confirm current specifics at source. Community repo luseloso/Factset_mcp_agent_example sourced from GitHub findings, user-owned, not an official FactSet release. Truthifi capability figures (semantic decoders, 130+ automated diagnostics, 18,000+ institutions, 99.7% accuracy, $6B+ in assets monitored) are company-reported, from truthifi.com / Truthifi platform data. Reviewed 2026-10-06. ↩↩↩↩

Related connectors in Advisor-Data-Analytics: Morningstar MCP · Get started with Truthifi · See how Truthifi keeps your accounts secure

Informational only — Truthifi provides data tools, not financial, investment, tax, or legal advice. Connecting your accounts gives your AI assistant read-only access to your own data; any decisions remain yours. Product details change — confirm current specifics on the provider's official site.

Truthifi is not an investment adviser and doesn't give investment, tax or legal advice.