Ask Claude what the crowd is saying: Quantral's MCP server is live

Quantral's investment ideas now reach the chat window where you already do the thinking. The Quantral MCP server is live, so Claude, ChatGPT, or any other MCP-compatible client can pull your Quantral data mid-conversation. Ask what investors are saying about a stock and the answer comes back scored and sourced, without opening another tab.

What shipped

The server gives a connected assistant five read-only things to ask for.

• The ranked list. The names with the strongest chatter right now, over a 24-hour or a 7-day window.
• One company's signal score. Both windows, each with its tier word: Exceptional, Strong, Moderate, Weak or Negative.
• The mentions behind a score. The individual posts a reading was built from, so you can read the case rather than take the number on trust.
• Monthly recaps. What the conversation around a company has been about, month by month.
• Company search, so the assistant knows which name you mean before it pulls anything.

In Claude and ChatGPT the answers arrive as interactive cards in the chat: a ranked list you can scan, a score card, a recap card, each name linking back to its page in the app.

What it looks like

This exchange ran on the morning of September 10.

The card came back with five names, all in the Strong tier: Nebius at 88, NVIDIA at 86, Micron at 84, Aehr Test Systems at 83 and SK Hynix at 83, each with a line on what the conversation was about. Nebius on a fresh price target, NVIDIA on paying up for AI chip test capacity, Micron and SK Hynix on analysts calling for memory prices to jump.

Micron came back at 84 over 24 hours and 84 over 7 days, Strong on both windows, with the current chatter pointing at that same memory pricing call. Two matching windows describe a conversation that held its level for the week instead of one that spiked overnight.

Then you keep pulling. Ask who is behind the Micron reading and how their past calls were graded, or whether the August recap told the same story, and the assistant chains the questions you would otherwise answer by hand across several screens. Every response carries the same framing, so a score stays a reading of the conversation and never becomes a forecast.

How to connect it

Setup takes about a minute and happens once. In Claude you add a custom connector under Settings; in ChatGPT you add it under Apps and connectors with developer mode on. Either way you sign in with your Quantral account and approve read access. Other MCP-compatible clients point at the same endpoint.

The /mcp page has the exact steps for each client, the endpoint URL, and a prompt you can paste into a fresh chat to have the assistant walk you through it. Agent access is included in every plan at no extra cost, and new accounts start with a 7-day free trial.

Who it is for

If you already think out loud in a chat window, your assistant now answers from the current reading instead of training data that went stale months ago, and you can push on that reading in plain language.

If you are catching up on a name you have not looked at since spring, the recaps do the heavy lifting: one question, several months of conversation, with the main voices named.

And if you build agents, the same endpoint is what a bot connects to. It is hosted, so there is nothing to run, authentication is an OAuth sign-in rather than an API key, and the limit is 60 requests a minute, which is well past what a scheduled strategy needs. Sentiment is one input among several; your strategy decides what weight it carries.

Try it

Open quantral.com/mcp to connect your assistant, and how to research stock sentiment with an AI assistant for the longer walkthrough, with the questions that do most of the work.

Scores above are a snapshot from the morning of September 10, 2026. For the current reading, ask your assistant, or open the Quantral app.

---

Quantral surfaces signals and context from public sources to support your own research. Nothing here is financial advice or a recommendation to buy or sell.