Horizon Trade says its no-code platform can turn plain-English ideas into backtested, broker-connected strategies, launching with $2 million in pre-seed funding and more than 23,000 people on a waitlist. Those figures establish financing and interest, not live performance, while the company’s own materials leave questions about deployment, pricing and marketplace readiness.
Horizon Trade says it has opened a shortcut from a trading idea to an automated order. The launch establishes that the startup has capital and attention; it does not yet establish how the system performs with customers’ money.
Horizon said in its July 16 announcement that the platform was launching with more than 23,000 traders and investors on its waitlist. A separate report on the company said Entrée Capital had invested $2 million in a pre-seed round.
Neither number is a usage metric. The available sources do not say how many people have paid for a plan, connected a broker, funded an account or completed a live trade. They also disclose no valuation or other financing terms, so the round’s significance beyond providing early operating capital cannot be measured from the retained evidence.
The company is unusually young. The report identifies founder and CEO Tuvia Ohana as 19 and says he finished high school at 16, previously bootstrapped another company and now leads a 15-person team whose average age is about 20. Those details explain part of the startup’s profile; they do not answer questions about financial-market infrastructure, security or execution.
The launch status itself needs sharper definition. A Horizon risk-management page published June 13 said the product was “being built” and directed readers to join a waitlist. Its current quickstart documentation says account registration is open. That sequence is consistent with a launch a month later, but it does not show how broadly broker-connected automation was available on day one.
Horizon’s pitch is straightforward: a user describes an asset, timeframe, entry and exit conditions, position sizing and risk limits in ordinary language. The system turns that description into explicit rules and code, tests it on historical data and prepares it for automated execution through a connected broker.
The company says it generates the underlying code, tests it across years of market data and stress-tests it for robustness. It also says the same strategy code runs in backtests and live trading, which can reduce errors introduced when logic is rewritten for deployment. Horizon claims support for Coinbase, Binance, Kraken, Alpaca, E*Trade, TradeStation and more than a dozen other venues in an account of the launch. Those are company-originated capability claims, not independently reported tests of the connections.
Identical code is only one layer of parity. Horizon’s quickstart warns that live results can depart from simulations because of fees, slippage, liquidity, latency, data quality and changing market conditions. It tells users to test different periods, model realistic costs where settings permit and avoid choosing a complex configuration only because it produced the best historical result. Available markets, order types and automation features also depend on the broker integration and the customer’s account.
The company’s FAQ reinforces the boundary: users can backtest without a broker, supported venues may change, and the customer remains responsible for the strategy, connection, permissions, risk controls and trades. The interface removes the need to write Python, Pine Script, MQL or broker API code; it does not remove the need to understand the generated rules.
The retained materials also disagree on the maturity of Horizon’s sharing feature. The launch announcement says users can copy portfolios and strategies shared by others, while the company report says Horizon “aims to create” a marketplace. None of the sources explains how shared strategies are vetted, how performance is presented or whether creators are paid.
Pricing is another blank. The FAQ tells prospective users to sign up for a plan, but the retained launch materials do not state plan prices, data entitlements, deployment limits or the total cost of using connected venues. That prevents a like-for-like economic comparison with established tools.
Horizon is not creating the category it is entering. Before Horizon’s launch, Composer already helped retail investors build and execute systematic strategies. SoFi acquired Composer and said customers would be able to express an idea in plain English, test it, automate it and draw from thousands of community-created strategies in one platform, according to reporting on the transaction.
SoFi did not disclose the acquisition price, but it reported 14.7 million members and $1.1 billion in adjusted revenue in the first quarter of 2026. That gives Composer a distribution channel and corporate base far larger than Horizon’s claimed waitlist. The same report noted that Robinhood had said it would let customers create dedicated accounts in which their own AI agents could trade stocks.
The products are not identical. Horizon emphasizes a research-to-execution layer connected to external equity and crypto venues; SoFi is integrating Composer into a broader financial platform; Robinhood’s announced approach lets customers bring agents to a brokerage account. Together, however, they narrow Horizon’s differentiation. Natural-language strategy building and automated execution are already competitive features. Horizon must distinguish itself through strategy coverage, data quality, broker reach, execution reliability or price—and its retained materials do not yet supply evidence for those comparisons.
Horizon describes itself as a technology and data provider, not a registered broker-dealer, investment adviser or fiduciary. Its risk guide says customers are solely responsible for trades executed through connected brokers. The platform can structure and automate rules, but its own instructions require users to review the logic, authorize account permissions, set exposure and loss limits, verify initial orders and keep monitoring after deployment.
That allocation of responsibility is not a finding that Horizon has violated a rule. FINRA’s general investor guidance does not name or assess Horizon. It warns, however, that auto-trading services from unregistered entities do not carry the oversight and investor protections required of registered firms, and it advises investors to verify claimed broker relationships, scrutinize unsupported performance or AI claims, protect account credentials and monitor even regulated services.
Horizon’s own warning captures the operational problem: automation can enforce sound rules consistently, but it can also execute bad rules faster. A polished backtest cannot show whether a generated strategy will remain valid in a new market regime or whether an integration will behave correctly during a data problem, delay or broker outage.
The central question will not be resolved by a larger waitlist. Horizon needs operating evidence: how many customers progress from prompts to paper trading and funded automation; which broker and exchange connections are live in each region; how modeled costs compare with actual fills; how often generated rules require correction; and how the system responds to stale data, rejected orders and interrupted connections.
Marketplace and business-model disclosures matter too. Users need to know whether shared-strategy results include costs, how overfitting and revisions are displayed, what creators earn, what each plan includes and which party handles a failed instruction.
Until those details arrive, Horizon’s clearest proposition is to reduce the programming required to test a rule-based idea. Whether it has also built dependable trading infrastructure remains the post-launch decision for customers—and the evidence the company still has to provide.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Anthropic’s 20-year lease gives TeraWulf a customer for 401 megawatts of future AI infrastructure, but rent starts only after delivery and the proposed utility deal passes power and grid costs to TeraWulf, leaving financing and project margin unresolved.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Twenty-nine countries signed an agreement creating WAICO as an independent intergovernmental organization, while China paired the launch with capacity-building offers that are not yet confirmed as WAICO programs.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
OpenAI says a handful of GPT-5.6 Sol file-deletion reports most commonly involved Full Access without sandboxing or Auto-review. Its own evaluations show a more complicated risk picture, and the company has not yet published an incident rate or evidence that its promised safeguards stop the failure.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
A rogue browser extension can trigger Claude for Chrome’s fixed Gmail, Docs, Calendar and business workflows; the attack is constrained by default approvals but can run silently for users who enabled unattended action.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Gold Eagle has begun collecting and prioritizing AI-discovered software vulnerabilities, but the voluntary federal clearinghouse has not disclosed results, operating rules or the resources that would turn findings into deployed fixes.
Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.