Best Brokers for AI Trading in 2026
Last updated June 2026
Short answer
The best broker for AI trading depends on how you want to use AI. For API-driven AI agents, Alpaca and Interactive Brokers expose the most open programmatic access, with Alpaca maintaining an official MCP server and a free paper-trading sandbox. For built-in, in-app AI, Public (Alpha) and Robinhood have the most developed native features, and SoFi adds an AI-flavored coach. Charles Schwab and Fidelity lead among full-service brokers with deep research and an automated-allocation option. If you would rather keep the broker you already use, an AI layer like Walnut connects most major US brokers through SnapTrade and lets you analyze and manage them by chatting through Claude or ChatGPT, read-only by default, with any trade you approve. Note that Robinhood is read-only for third-party tools; Public, Alpaca, Schwab, Tradier, and Webull support trading.
“Best broker for AI trading” sounds like one ranking, but it splits into three different questions. Do you want to write code so an AI agent trades programmatically? Do you want AI features built into the broker's own app? Or do you want to keep your current broker and add an AI layer on top? This guide covers eight brokers (Interactive Brokers, Alpaca, Charles Schwab, Fidelity, Robinhood, Webull, Public, and SoFi) on the same fields, ranks them by use-case, and explains where Walnut fits as the AI layer you run over a broker rather than a broker itself. We are honest about where each one, including Walnut, is the wrong fit.
The three ways a broker meets AI trading
Before comparing brokers, it helps to name the three patterns, because a broker that is great for one is often weak for another.
- API-driven AI agents. The broker exposes a programmatic API (and sometimes an MCP server) so a bot or LLM agent reads data and places orders directly. Alpaca and Interactive Brokers lead here. This is the most powerful and the most technical path.
- Built-in broker AI. The AI lives inside the broker's own app. Public's Alpha assistant and Robinhood's Cortex are examples. It is the easiest to use, but the AI is confined to that app and that broker.
- An AI layer on top of any broker. A separate tool connects the broker you already have (usually through a regulated aggregator like SnapTrade) and adds AI analysis and chat across it. Walnut is this kind. You keep your broker; the layer reads it and, where supported, trades it with your approval.
The right broker is the one that matches the pattern you actually want. The takeaway throughout this guide: there is no single overall winner, only a best fit per approach.
What makes a broker good for AI trading
When a broker claims to support AI trading, these are the specific things worth checking. A strong fit covers the ones that match your approach; a weak fit only gestures at AI in marketing.
- Open API or MCP access. Whether an outside agent can read data and place trades programmatically. Alpaca offers an official MCP server; IBKR has documented trading APIs. Most consumer apps do not expose retail trading APIs.
- Third-party trade permission. Whether a connected tool can actually place orders, not just read. Through SnapTrade, Public, Alpaca, Schwab, Tradier, and Webull support trading, while Robinhood is read-only.
- Built-in AI quality. If you prefer in-app AI, whether the assistant answers real questions about stocks and your account (Public's Alpha, Robinhood's Cortex) rather than restating a help article.
- Fractional shares. Essential when an AI builds a multi-stock basket from a fixed dollar amount. Most brokers here support fractional shares; Public adds a per-order minimum.
- Cost. Commissions (mostly zero on US stocks here), margin rates, premium AI tiers (like Robinhood Gold), and fund expenses. The headline price is rarely the whole cost.
- Read-only vs trade-enabled, and approval. Whether an AI can ever place an order without you confirming it. Autonomous bots and approval-gated assistants are different risk profiles.
- Security and data handling. The safer connections never store your broker login and default to read-only access through a regulated aggregator such as SnapTrade.
The eight brokers worth knowing for AI trading
Each broker below is described on the same six fields, so you can scan across them: what it is, its AI features, how an API or AI assistant can access it, fractional support and cost, who it suits, and one honest limitation. Walnut follows as the AI layer you run on top of these brokers rather than a broker itself.
Interactive Brokers (IBKR)
A professional-grade broker with the broadest market access in the category: US and international stocks, options, futures, bonds, and currencies, with low margin rates and deep order types.
- AI features: Documented trading APIs (the TWS API and the Client Portal Web API) that AI agents can drive, plus community-built MCP servers; its research tooling has added AI-assisted features over time.
- API or AI-assistant access: Full programmatic API for read and trade; AI-assistant access is mostly via the API or community-built MCP connectors rather than a polished built-in chatbot.
- Fractional and cost: Supports fractional shares; very low commissions and margin rates, though the fee schedule is more complex than the zero-commission consumer apps.
- Best for: Advanced and international investors who want programmatic, API-driven AI trading with the widest instrument coverage.
- One honest limitation: Powerful but complex, and the official AI-assistant story is thinner than the API story, so most AI integrations are community-built.
Alpaca
A developer-first, commission-free brokerage built API-first, designed for programmatic trading rather than a consumer trading app. It is the broker most often wired directly into trading bots and AI agents.
- AI features: An official Alpaca MCP server that gives AI agents real-time market data, paper trading, and live execution; rich REST and streaming APIs that LLM-based agents call directly.
- API or AI-assistant access: Full API for read, trade, and paper trading, plus a maintained official MCP server, so connecting an AI agent is first-class rather than an afterthought.
- Fractional and cost: Supports fractional shares; commission-free on US stocks and ETFs, with a free paper-trading environment to test agents safely.
- Best for: Developers and quant-minded investors building AI agents that read data and place trades programmatically.
- One honest limitation: Developer-first: it expects you to write code or run an MCP server, so it is not a point-and-click app for non-technical users.
Charles Schwab
A large, full-service US broker (which absorbed TD Ameritrade) offering stocks, ETFs, options, and mutual funds, with deep research, banking, and a robo-advisor (Schwab Intelligent Portfolios).
- AI features: Schwab Intelligent Portfolios automates a diversified allocation; the platform offers extensive research and screeners, and Schwab has trading APIs (the former TD Ameritrade API lineage) that third-party tools and agents can use.
- API or AI-assistant access: Trading API access for approved developers and third-party apps; can be connected to AI assistants through aggregators like SnapTrade, with trading supported.
- Fractional and cost: Supports fractional shares of S&P 500 stocks (Schwab Stock Slices); commission-free US stock and ETF trades.
- Best for: Investors who want a large, established broker with research depth plus an automated-allocation option and API access.
- One honest limitation: Its built-in AI is more automated-allocation and research tooling than a conversational assistant over your specific holdings.
Fidelity
A large, full-service US broker known for low-cost index funds (including zero-expense-ratio funds), strong research, and a wide product range across brokerage, retirement, and cash management.
- AI features: Deep research, screeners, and planning tools, and a robo-advisor (Fidelity Go); Fidelity has experimented with AI-assisted research and customer features, but it does not market an open trading API for retail AI agents the way Alpaca or IBKR do.
- API or AI-assistant access: Connectable to third-party AI tools through aggregators like SnapTrade for tracking and analysis; native retail trading-API access for outside AI agents is limited.
- Fractional and cost: Supports fractional shares (Stocks by the Slice); commission-free US stock and ETF trades and notably low-cost funds.
- Best for: Long-term investors who want low-cost funds, strong research, and an automated option, and who connect AI tools mainly for analysis.
- One honest limitation: Less open as a programmatic trading platform for AI agents, so AI use leans toward analysis rather than agent-driven execution.
Robinhood
A mobile-first, commission-free broker that popularized fractional shares and simple investing, with stocks, ETFs, options, and crypto and a paid Gold tier.
- AI features: Has rolled out AI-assisted features inside its own app (including Cortex for Gold subscribers, aimed at market and stock context); the experience lives inside Robinhood rather than being open to outside agents.
- API or AI-assistant access: Read-only through the major third-party aggregators: via SnapTrade, Robinhood connects for tracking and analysis only, with no third-party trade execution. Trades happen inside Robinhood's own app.
- Fractional and cost: Supports fractional shares; commission-free trading, with the deepest AI features reserved for the paid Gold tier.
- Best for: Mobile-first investors who want a simple, commission-free app and are fine using AI features inside Robinhood itself.
- One honest limitation: Read-only via SnapTrade for third-party tools, so an outside AI assistant can analyze a Robinhood account but cannot place trades in it.
Webull
A commission-free broker with advanced charting and an active-trader feel, covering US stocks, ETFs, options, and (in many regions) crypto, popular with more hands-on traders.
- AI features: Advanced charting, technical indicators, and screening that suit active trading; Webull has added AI-assisted market features in-app, and it is one of the brokers where third-party tools can place trades.
- API or AI-assistant access: Connectable through aggregators like SnapTrade with trading supported, so an AI layer on top can both read and (with your approval) trade a Webull account.
- Fractional and cost: Supports fractional shares; commission-free US stock and ETF trades, with a more trading-oriented interface than the beginner apps.
- Best for: Active traders who want advanced charts and a broker that third-party AI tools can actually trade in.
- One honest limitation: The interface and feature set lean toward active trading, which can be more than a hands-off, long-term investor needs.
Public
A modern investing app for stocks, ETFs, bonds, and a high-yield cash option, with a clean interface and a focus on transparency, including AI-assisted context on companies inside the app.
- AI features: An in-app AI assistant (Alpha) that answers questions about stocks and the market in plain language; the product wraps that around a straightforward investing experience.
- API or AI-assistant access: Connectable through aggregators like SnapTrade with trading supported, so an external AI layer can read and place trades. Note that Public enforces a per-order minimum on fractional trades.
- Fractional and cost: Supports fractional shares; commission-free equity trades. Fractional orders carry a small per-order minimum, so very tiny single-stock buys are not possible.
- Best for: Investors who want a clean modern app with in-app AI context and the ability to be traded by an outside AI layer.
- One honest limitation: The per-order fractional minimum means each leg of a many-stock basket has to clear that floor, which matters for small accounts.
SoFi (SoFi Invest)
An all-in-one money app bundling brokerage, automated investing, banking, and lending, aimed at beginners who want investing alongside the rest of their finances.
- AI features: Automated investing plus an AI-flavored financial coach inside the app; the AI is oriented toward guidance and the broader money product rather than open agent access.
- API or AI-assistant access: Primarily an in-app experience; third-party AI-agent and trading-API access is limited compared with Alpaca or IBKR.
- Fractional and cost: Supports fractional shares; commission-free stock and ETF trades and free automated investing.
- Best for: Beginners who want investing, banking, and automated portfolios in one app with light AI guidance.
- One honest limitation: Lighter on deep research and on open access for outside AI agents, so it suits in-app use more than programmatic AI trading.
Walnut (the AI layer on top)
Not a broker. Walnut connects the brokerage you already have through SnapTrade and adds the AI layer: you analyze and manage what you hold by talking through Claude, ChatGPT, or a built-in assistant, and trades route back through your own broker.
- AI features: Conversational analysis of your real holdings, each framed against the S&P 500 (outperforming, in line, or lagging), plus thematic baskets you can build around a thesis and invest into.
- API or AI-assistant access: Sits on top of most major US brokers via SnapTrade. Read-only by default; it can place trades only with your approval, and only at brokers SnapTrade supports for trading (Public, Alpaca, Schwab, Tradier, Webull). Robinhood connects read-only.
- Fractional and cost: Uses your broker's fractional support and pricing rather than charging brokerage commissions itself; has a free tier.
- Best for: People who want an AI layer across whatever broker they already use, instead of switching brokers to get AI features.
- One honest limitation: It is a layer, not a broker, so you must connect an existing account, and broker feeds rarely pass cost basis (returns are framed as window returns, not realized profit and loss).
At a glance
| Broker | AI / third-party access | Fractional & cost | Best for |
|---|---|---|---|
| Interactive Brokers (IBKR) | Full programmatic API for read and trade; AI-assistant access is mostly via the API or community-built MCP connectors rather than a polished built-in chatbot | Supports fractional shares; very low commissions and margin rates, though the fee schedule is more complex than the zero-commission consumer apps | Advanced and international investors who want programmatic, API-driven AI trading with the widest instrument coverage |
| Alpaca | Full API for read, trade, and paper trading, plus a maintained official MCP server, so connecting an AI agent is first-class rather than an afterthought | Supports fractional shares; commission-free on US stocks and ETFs, with a free paper-trading environment to test agents safely | Developers and quant-minded investors building AI agents that read data and place trades programmatically |
| Charles Schwab | Trading API access for approved developers and third-party apps; can be connected to AI assistants through aggregators like SnapTrade, with trading supported | Supports fractional shares of S&P 500 stocks (Schwab Stock Slices); commission-free US stock and ETF trades | Investors who want a large, established broker with research depth plus an automated-allocation option and API access |
| Fidelity | Connectable to third-party AI tools through aggregators like SnapTrade for tracking and analysis; native retail trading-API access for outside AI agents is limited | Supports fractional shares (Stocks by the Slice); commission-free US stock and ETF trades and notably low-cost funds | Long-term investors who want low-cost funds, strong research, and an automated option, and who connect AI tools mainly for analysis |
| Robinhood | Read-only through the major third-party aggregators: via SnapTrade, Robinhood connects for tracking and analysis only, with no third-party trade execution. Trades happen inside Robinhood's own app | Supports fractional shares; commission-free trading, with the deepest AI features reserved for the paid Gold tier | Mobile-first investors who want a simple, commission-free app and are fine using AI features inside Robinhood itself |
| Webull | Connectable through aggregators like SnapTrade with trading supported, so an AI layer on top can both read and (with your approval) trade a Webull account | Supports fractional shares; commission-free US stock and ETF trades, with a more trading-oriented interface than the beginner apps | Active traders who want advanced charts and a broker that third-party AI tools can actually trade in |
| Public | Connectable through aggregators like SnapTrade with trading supported, so an external AI layer can read and place trades. Note that Public enforces a per-order minimum on fractional trades | Supports fractional shares; commission-free equity trades. Fractional orders carry a small per-order minimum, so very tiny single-stock buys are not possible | Investors who want a clean modern app with in-app AI context and the ability to be traded by an outside AI layer |
| SoFi (SoFi Invest) | Primarily an in-app experience; third-party AI-agent and trading-API access is limited compared with Alpaca or IBKR | Supports fractional shares; commission-free stock and ETF trades and free automated investing | Beginners who want investing, banking, and automated portfolios in one app with light AI guidance |
| Walnut (the AI layer on top) | Sits on top of most major US brokers via SnapTrade. Read-only by default; it can place trades only with your approval, and only at brokers SnapTrade supports for trading (Public, Alpaca, Schwab, Tradier, Webull). Robinhood connects read-only | Uses your broker's fractional support and pricing rather than charging brokerage commissions itself; has a free tier | People who want an AI layer across whatever broker they already use, instead of switching brokers to get AI features |
Ranked by how you want to use AI
There is no overall number one, because the right broker depends on whether you want to code an agent, use in-app AI, add a layer on top, or get a full-service broker with research. Below the field is ranked inside each use-case, with the stronger fit first. Walnut leads only in its own category (an AI layer on top of any broker), not across the board.
Best for API-driven, agent-built AI trading
If you want to write code or run an MCP server so an AI agent reads data and places trades directly, these expose the most open programmatic access.
- 1. Alpaca. API-first and commission-free, with an official MCP server for real-time data, paper trading, and live execution, so wiring up an AI agent is first-class.
- 2. Interactive Brokers (IBKR). Documented trading APIs and the broadest instrument coverage; AI connectors are mostly community-built but the programmatic control is deep.
Best for built-in, in-app broker AI
If you would rather use AI features inside the broker's own app than connect an outside tool, these have the most developed native AI.
- 1. Public. An in-app AI assistant (Alpha) answers plain-language questions about stocks and the market right inside a clean investing app.
- 2. Robinhood. Rolling out in-app AI features (Cortex for Gold subscribers) for market and stock context, kept inside Robinhood's own experience.
Best for an AI layer on top of any broker
If you want to keep the broker you already use and add AI by connecting it to an assistant, the question is which layer reads (and, where supported, trades) your account. Walnut leads here, but only here.
- 1. Walnut (the AI layer on top). Connects most major US brokers through SnapTrade and lets you analyze and manage them by chatting through Claude or ChatGPT, with each holding framed against the S&P 500. Read-only by default; you approve any trade.
- 2. Alpaca. If you are technical, Alpaca's official MCP server is the cleanest direct path to put an AI agent on a brokerage account, though it is a separate account rather than a layer over your existing one.
Best full-service broker with research and an automated option
If you want a large, established broker with deep research and the option of automated allocation, with AI used mainly for analysis, these lead.
- 1. Charles Schwab. Research depth, Schwab Intelligent Portfolios for automated allocation, and trading-API access that third-party tools can use.
- 2. Fidelity. Low-cost funds, strong research, and Fidelity Go for automation; best connected to AI tools for analysis rather than agent-driven trading.
Which brokers have built-in AI?
A handful build AI directly into their own apps, which is the easiest way to use it but keeps the AI confined to that broker. Public has an in-app assistant called Alpha that answers plain-language questions about stocks and the market. Robinhood has rolled out AI features including Cortex for its paid Gold tier, oriented toward market and stock context. SoFi pairs automated investing with an AI-flavored financial coach inside its all-in-one money app. Charles Schwab and Fidelity lean more on deep research tools and robo-advisors (Schwab Intelligent Portfolios and Fidelity Go) than on a conversational chatbot over your specific holdings. If you want AI without leaving your broker, Public and Robinhood have the most visible native features today.
Can I connect a broker to an AI assistant?
Yes, and this is the path that does not require switching brokers or writing code. On their own, Claude and ChatGPT cannot see your holdings, so they answer with generic, made-up examples. Connecting your brokerage through a tool like Walnut, which uses the regulated SnapTrade aggregator, gives the assistant read access to your real positions, so the analysis is about what you actually own: your concentration, your performance versus the S&P 500, what has changed. Through SnapTrade, trading is supported on Public, Alpaca, Schwab, Tradier, and Webull, so the assistant can also place orders you approve, while Robinhood connects read-only (analysis only). Developers have a second route: Alpaca's official MCP server and Interactive Brokers' APIs let you wire an AI agent directly into a brokerage account in code.
Which broker is best for AI trading?
The honest answer is that it splits by approach, so the “best” broker is the one matching how you want to use AI. For building an AI agent that trades programmatically, Alpaca is the strongest fit: it is API-first, commission-free, and ships an official MCP server plus a free paper-trading sandbox to test agents safely. Interactive Brokers offers the broadest market access with documented trading APIs, though its AI connectors are mostly community-built. For in-app AI without any setup, Public and Robinhood lead. For a full-service broker with research and an automated option, Charles Schwab and Fidelity lead. And if you would rather keep the broker you already use and simply add AI on top, an AI layer like Walnut connects most major US brokers through SnapTrade. There is no single winner; match the broker to whether you want to code, use in-app AI, get research, or add a layer.
How we evaluated these
We limited the field to brokers (plus the layer that sits on top of them) that have a real AI-trading angle, whether that is an API for agents, native in-app AI, or third-party connectability. Within that set we weighed a few things specific to AI trading:
- Access model: whether an outside AI agent can read and trade via an API or MCP server, and whether third-party tools can place orders at all.
- Built-in AI depth: how far the broker's own assistant goes past marketing into real questions about stocks and your account.
- Trade permissions and approval: read-only vs trade-enabled through aggregators, and whether anything can trade without your confirmation.
- Fractional support and total cost: commissions, margin, premium AI tiers, fund expenses, and per-order minimums.
- Honesty of the marketing: we marked down anything implying AI guarantees market-beating returns, because no broker or tool can promise that.
We did not crown a single overall winner. The best broker depends on how you want to use AI and how hands-on you are. Features, pricing, and supported-broker lists change; treat the specifics here as a starting point and verify on each provider's site.
Where Walnut fits
To be upfront, since this is our site: Walnut is not a broker. It is the AI layer you run on top of the broker you already use, and it leads only in that category, not overall. It connects most major US brokers through SnapTrade and lets you analyze and manage what you hold by talking through Claude, ChatGPT, or a built-in assistant. Its dashboard frames each holding's return against the S&P 500 and classifies it as outperforming, in line, or lagging, alongside momentum and concentration reads, and you can build thematic baskets around a thesis and invest into them. Because broker feeds rarely pass cost basis, Walnut uses a window-return framing rather than realized profit and loss, and says so. It is read-only by default, every trade needs your approval, and it can place trades only at SnapTrade-supported brokers (Public, Alpaca, Schwab, Tradier, Webull); Robinhood connects read-only. You keep the broker you already have. Walnut is not an investment adviser.
Where Walnut is the wrong choice
Just as importantly, here is when going to a broker directly fits the job better:
- You want to build an AI agent that trades programmatically. Alpaca is API-first with an official MCP server and paper trading; Walnut is a no-code layer, not a developer platform.
- You want the broadest instrument coverage with deep programmatic control. Interactive Brokers spans global stocks, options, futures, and bonds with documented APIs; Walnut focuses on US equities and ETFs through your broker.
- You want AI entirely inside one app and do not want to connect anything. Public's Alpha or Robinhood's Cortex keep the AI in-app; Walnut needs you to connect an existing brokerage.
- You do not have a brokerage account yet. Walnut sits on top of one, so you would open an account at a broker like Public, Schwab, Fidelity, or Alpaca first.
- You want a robo-advisor to run your money for you. Schwab Intelligent Portfolios, Fidelity Go, or SoFi's automated investing manage a portfolio; Walnut keeps you in control with approval on every trade.
- Your only broker is Robinhood and you want an outside tool to trade it. Robinhood is read-only via SnapTrade, so a layer can analyze but not trade it; use a SnapTrade-trading broker for that.
From a connected account you can dig into a specific stock, an ETF you hold, or a theme you want exposure to. For the wider field, see the best AI investing apps roundup, or how to connect your brokerage to an AI assistant.
Try Walnut on top of your broker
Walnut connects most major US brokers in a few clicks, then analyzes what you hold against the S&P 500 and lets you ask questions through Claude, ChatGPT, or its built-in AI. Read-only by default; you approve every trade.
FAQ
What is the best broker for AI trading in 2026?
+
There is no single best one; it depends on how you want to use AI. For API-driven AI agents, Alpaca and Interactive Brokers expose the most open programmatic access. For built-in in-app AI, Public (Alpha) and Robinhood have developed native features. If you want to keep your current broker and add an AI layer on top, a tool like Walnut connects most major US brokers through SnapTrade and lets you analyze and manage them by chat.
Which brokers have built-in AI?
+
Several layer AI inside their own apps. Public has an in-app assistant (Alpha) for plain-language questions about stocks and the market. Robinhood has rolled out AI features including Cortex for its Gold tier. SoFi pairs automated investing with an AI-flavored coach. Schwab and Fidelity lean more on research tools and robo-advisors (Intelligent Portfolios, Fidelity Go) than on a conversational chatbot over your holdings.
Can I connect a broker to an AI assistant like Claude or ChatGPT?
+
Yes. On their own, Claude and ChatGPT cannot see your holdings, so they give generic answers. Connecting your brokerage through a tool like Walnut (which uses the SnapTrade aggregator) gives the assistant read access to your real positions, so the analysis is about what you actually own. Alpaca and Interactive Brokers also let developers wire agents in via APIs or MCP servers.
Which broker is best for AI trading?
+
It splits by approach. Alpaca is the strongest API-first choice for building an AI agent that trades programmatically, with an official MCP server and a free paper-trading sandbox. Interactive Brokers offers the broadest market access with documented APIs. If you do not want to code or switch brokers, an AI layer like Walnut sits on top of the broker you already use. Match the broker to whether you want to code, use in-app AI, or add a layer.
Does Alpaca support AI trading?
+
Yes, more directly than most. Alpaca is a developer-first, commission-free broker built API-first, and it maintains an official MCP server that gives AI agents real-time data, paper trading, and live execution. That makes it the broker most often wired straight into trading bots and AI agents. The trade-off is that it expects you to write code or run a server, so it is not a point-and-click app.
Can I trade my Robinhood account with an AI tool?
+
Not through third-party tools. Robinhood connects read-only through the major aggregator SnapTrade, which means an outside AI assistant can analyze a Robinhood account but cannot place trades in it. Robinhood's own in-app AI features (like Cortex for Gold) work inside its app. To let an external AI layer place trades, use a broker SnapTrade supports for trading, such as Public, Alpaca, Schwab, Tradier, or Webull.
Which brokers can a third-party AI tool actually place trades in?
+
Through the SnapTrade aggregator, trading is supported on brokers including Public, Alpaca, Schwab, Tradier, and Webull, so an AI layer on top can read and (with your approval) place orders. Robinhood is read-only via SnapTrade, meaning analysis only. Access changes over time, so verify the current supported-broker and trading list on the provider's site before relying on it.
Do I have to switch brokers to use AI for investing?
+
No. That is the point of an AI layer. Brokers like Public and Robinhood build AI into their own apps, and Alpaca and IBKR expose APIs for agents, but you do not have to move accounts to get AI. A tool like Walnut connects the broker you already have through SnapTrade and adds conversational analysis on top, read-only by default, so you keep your existing account.
What is the cheapest broker for AI trading?
+
Most consumer brokers here are commission-free on US stocks and ETFs (Alpaca, Schwab, Fidelity, Robinhood, Webull, Public, SoFi), so cost differences are usually in margin rates, premium tiers, and fund expenses rather than per-trade commissions. Alpaca is free and includes a paper-trading sandbox for testing agents. Watch for paid AI tiers (such as Robinhood Gold) and any per-order fractional minimums (such as Public's).
Do these brokers support fractional shares for AI-built baskets?
+
Most do. Alpaca, Schwab (Stock Slices), Fidelity (Stocks by the Slice), Robinhood, Webull, Public, and SoFi all support fractional shares, which matters when an AI builds a multi-stock basket from a fixed dollar amount. Note that Public applies a per-order minimum on fractional trades, so each leg of a basket has to clear that floor, which can constrain very small accounts.
Is an AI trading bot the same as an AI investing assistant?
+
No. An AI trading bot runs an automated strategy and places trades on its own, which suits API-first brokers like Alpaca. An AI investing assistant is conversational: you ask questions about your holdings and it explains, suggests, or (with approval) helps you act. Walnut is the assistant kind, read-only by default, with any trade requiring your approval, rather than an autonomous bot.
How does Walnut work with these brokers?
+
Walnut is not a broker; it is the AI layer you run on top of one. It connects most major US brokers through SnapTrade, reads your holdings, and lets you analyze and manage them by talking through Claude, ChatGPT, or a built-in assistant, with each holding framed against the S&P 500. It is read-only by default and places trades only with your approval, and only at SnapTrade-supported brokers. Walnut is not an investment adviser.
Walnut is informational and is not an investment adviser. Broker features, pricing, supported-broker lists, and availability change; verify current details on each provider's site before deciding. Nothing on this page is a recommendation to buy, sell, or hold any security or to use any particular broker or product.