Introducing

Build on Massive.com with llms.txt

Jan 22, 2026

In this post, we'll dive into our newly implemented llms.txt support in our API documentation. The llms.txt format enables AI assistants to retrieve documentation directly in structured markdown, reducing the need for manual copying from HTML pages. Below, we cover the format, usage with AI tools, and access patterns.

The llms.txt Format

The llms.txt specification defines a standard for serving documentation at predictable URLs in markdown, allowing AI systems to fetch and parse content without HTML processing or scraping. It mirrors robots.txt in structure but focuses on accessibility for language models.

We have implemented this across all API documentation. The primary index at https://massive.com/docs/rest/llms.txt lists endpoints by category, with links to detailed markdown files. AI tools can use this index to select and retrieve specific endpoint docs as needed, optimizing context usage.

Usage with AI Coding Assistants

If you're using AI tools with web-fetch capabilities, such as Claude or Cursor, you can start with the REST index:

https://massive.com/docs/rest/llms.txt

This returns a structured index of every REST endpoint we offer, organized by category (stocks, crypto, forex, options, and more). Each entry links to the full documentation for that endpoint. Your AI assistant can read the index, identify which endpoints are relevant to your task, and fetch the detailed docs it needs.

You can access these resources via the following URLs:

• All REST endpoints index: https://massive.com/docs/rest/llms.txt
• Section index (e.g., stocks): https://massive.com/docs/rest/stocks/llms.txt
• Full section docs: https://massive.com/docs/rest/stocks/llms-full.txt
• Single endpoint: https://massive.com/docs/rest/stocks/news.md
• WebSocket index: https://massive.com/docs/websocket/llms.txt
• WebSocket (full sections): https://massive.com/docs/websocket/llms-full.txt

Here's an example:

# REST API

## Stocks

- [Custom Bars (OHLC)](https://massive.com/docs/rest/stocks/aggregates/custom-bars.md): Retrieve aggregated historical OHLC (Open, High, Low, Close) and volume data for a specified stock ticker over a custom date range and time interval in Eastern Time (ET)
- [Daily Market Summary (OHLC)](https://massive.com/docs/rest/stocks/aggregates/daily-market-summary.md): Retrieve daily OHLC (open, high, low, close), volume, and volume-weighted average price (VWAP) data for all U.S
- [Daily Ticker Summary (OHLC)](https://massive.com/docs/rest/stocks/aggregates/daily-ticker-summary.md): Retrieve the opening and closing prices for a specific stock ticker on a given date, along with any pre-market and after-hours trade prices
...

The AI reads the index, identifies relevant ones (e.g., aggregates for building a stock screener or options flow for derivatives analysis), and fetches only what's needed, saving on context tokens and costs.

You can also narrow it down by asset class for focused queries:

• Stocks: https://massive.com/docs/rest/stocks/llms.txt
• Crypto: https://massive.com/docs/rest/crypto/llms.txt
• Forex: https://massive.com/docs/rest/forex/llms.txt
• Options: https://massive.com/docs/rest/options/llms.txt
• WebSockets: https://massive.com/docs/websocket/llms.txt

For instance, prompting with the stocks index lets your AI zero in on equities without clutter from other categories.

Usage in Chat Interfaces

For chat-based AIs without automatic fetching, like ChatGPT, use llms-full.txt variants via https://massive.com/docs/rest/stocks/llms-full.txt.

This endpoint returns complete markdown for all endpoints in the section within the response. Just copy the output and include it in the chat prompt for full context on parameters, responses, and schemas.

# REST API

## Aggregates

# REST
## Stocks

### Custom Bars (OHLC)

**Endpoint:** `GET /v2/aggs/ticker/{stocksTicker}/range/{multiplier}/{timespan}/{from}/{to}`

**Description:**

Retrieve aggregated historical OHLC (Open, High, Low, Close) and volume data for a specified stock ticker over a custom date range and time interval in Eastern Time (ET). Aggregates are constructed exclusively from qualifying trades that meet specific conditions. If no eligible trades occur within a given timeframe, no aggregate bar is produced, resulting in an empty interval that indicates a lack of trading activity during that period. Users can tailor their data by adjusting the multiplier and timespan parameters (e.g., a 5-minute bar), covering pre-market, regular market, and after-hours sessions. This flexibility supports a broad range of analytical and visualization needs.

Use Cases: Data visualization, technical analysis, backtesting strategies, market research.

## Path Parameters

| Parameter | Type | Required | Description |
| --- | --- | --- | --- |
| `stocksTicker` | string | Yes | Specify a case-sensitive ticker symbol. For example, AAPL represents Apple Inc. |
| `multiplier` | integer | Yes | The size of the timespan multiplier. |
| `timespan` | string | Yes | The size of the time window. |
| `from` | string | Yes | The start of the aggregate time window. Either a date with the format YYYY-MM-DD or a millisecond timestamp. |
| `to` | string | Yes | The end of the aggregate time window. Either a date with the format YYYY-MM-DD or a millisecond timestamp. |
...

This approach supports queries like identifying endpoints for stocks quotes, providing details on all available options. For individual endpoints, you can now append

to the URL, like https://massive.com/docs/rest/stocks/news.md; alternatively, use the "Open as Markdown" button on the HTML page. This yields markdown with endpoint specifics, including endpoint paths and methods, parameter details such as types, defaults, constraints, and operators (e.g., , , date ranges), response schemas with field definitions, and sample JSON responses. This matches the HTML documentation in detail, formatted for parsing.

Next Steps

We hope you find the llms.txt support across our API documentation useful, as it provides a standardized way for AI assistants to access structured markdown files directly, bypassing manual HTML copying and pasting.

To apply this in your projects, test the llms.txt endpoints with your preferred AI tool, starting with scoped indexes for specific asset classes.

• 

  Rian Dolphin

  rian

  Read More