# Ornn Data

> Ornn Data: GPU and Memory Market Data
The Ornn Data API is a read-only REST API for real-time and historical GPU and memory price indices.

## Docs

- [Analytics](https://data.ornn.com/docs/analytics.md): Annualized volatility and market utilization derived from the GPU price index.
- [Get annualized price volatility](https://data.ornn.com/docs/api-reference/analytics/get-annualized-price-volatility.md): Standard deviation of daily log returns over a rolling window, annualized. Results are most-recent-first.
- [Get market utilization](https://data.ornn.com/docs/api-reference/analytics/get-market-utilization.md): The share of tracked capacity that is rented, between 0 and 1.
- [Get current price for one GPU](https://data.ornn.com/docs/api-reference/current-prices/get-current-price-for-one-gpu.md): The current price index for a single GPU model, in USD per GPU-hour.
- [Get latest daily index for all GPUs](https://data.ornn.com/docs/api-reference/current-prices/get-latest-daily-index-for-all-gpus.md): The latest daily index for every tracked GPU in one call. Public, no key required.
- [Get latest daily index for one GPU](https://data.ornn.com/docs/api-reference/current-prices/get-latest-daily-index-for-one-gpu.md): The latest daily index value for one GPU. Public, no key required.
- [Get previous day's daily index](https://data.ornn.com/docs/api-reference/current-prices/get-previous-days-daily-index.md): The previous day's daily index for a GPU. Useful for day-over-day comparisons.
- [Get today's daily index](https://data.ornn.com/docs/api-reference/current-prices/get-todays-daily-index.md): **Deprecated — use `/api/daily-index` instead**, which returns the same most-recent daily index value for one GPU. This alias is kept for backward compatibility and may be removed in a future version. For the current (hourly) price use `/api/gpu/{gpuName}`; for every GPU at once use `/api/daily-inde…
- [Get a clean daily/hourly series](https://data.ornn.com/docs/api-reference/historical-prices/get-a-clean-dailyhourly-series.md): A clean timestamp / index_value series, ideal for charts and DataFrames. Points are most-recent-first.
- [Get a public daily series](https://data.ornn.com/docs/api-reference/historical-prices/get-a-public-daily-series.md): Daily index series for a public GPU. Without a key, the free tier covers 5 GPUs — `H100 SXM`, `H200`, `B200`, `A100 SXM4`, `RTX 5090` — for the trailing 3 months (the same window charted for free on index.ornn.com); any other GPU (e.g. `RTX PRO 6000 WS`) requires a key. Pass your API key to unlock t…
- [Get history rows with metadata](https://data.ornn.com/docs/api-reference/historical-prices/get-history-rows-with-metadata.md): Richer history rows over a date range, with per-row granularity.
- [Get index value nearest a timestamp](https://data.ornn.com/docs/api-reference/historical-prices/get-index-value-nearest-a-timestamp.md): Returns the index value of the record nearest the requested time, with the gap reported in time_difference.
- [Get the full H100 daily series](https://data.ornn.com/docs/api-reference/historical-prices/get-the-full-h100-daily-series.md): The H100 SXM global daily index as a clean timestamp and value series. Without a key you get the trailing 3 months (the same free window on index.ornn.com); pass your API key to unlock the full history back to the earliest available data. The response includes `access: "public-3mo" | "full"`.
- [Get the daily history for an LLM coding product](https://data.ornn.com/docs/api-reference/llm-coding/get-the-daily-history-for-an-llm-coding-product.md): Daily history for an LLM coding-activity index, oldest first. Requires an API key — LLM coding analytics have no free tier. Note this is a coding-activity index, not one of the public token price indices (those are OTPI, under `/api/otpi`). Single-tool products return rows with a single `value`; com…
- [Get the latest value for an LLM coding product](https://data.ornn.com/docs/api-reference/llm-coding/get-the-latest-value-for-an-llm-coding-product.md): The most recent daily value for a single (non-compare) LLM coding product, with the day-over-day percent change. Requires an API key — LLM coding analytics have no free tier. `changePct` is `null` when there is no previous day on file or the previous value was zero. Compare products such as `cc-vs-c…
- [List available LLM coding products](https://data.ornn.com/docs/api-reference/llm-coding/list-available-llm-coding-products.md): The list of pre-built products exposed by the LLM coding endpoints. Each product pairs a tool with a metric (PRs, commits, or merge rate). Use this to drive a selector without hard-coding product IDs.
- [Get current memory spot prices](https://data.ornn.com/docs/api-reference/memory/get-current-memory-spot-prices.md): Current memory spot prices. With no params, returns all types. Narrow with type or category.
- [Get memory price history](https://data.ornn.com/docs/api-reference/memory/get-memory-price-history.md)
- [List tracked memory types](https://data.ornn.com/docs/api-reference/memory/list-tracked-memory-types.md)
- [Aggregate sites by country or GPU type](https://data.ornn.com/docs/api-reference/neo-cloud-sites/aggregate-sites-by-country-or-gpu-type.md): Aggregate power capacity, GPU quantity, and site counts grouped by country or by GPU type. Useful for regional/technology risk rollups. Defaults to neo-cloud sites; `scope=all` widens to the full set. For `by=gpu_type`, a site listing multiple chips (e.g. `H100/A100`) contributes to each of its chip…
- [Query neo-cloud sites by region](https://data.ornn.com/docs/api-reference/neo-cloud-sites/query-neo-cloud-sites-by-region.md): Return GPU data-center sites in a geographic region, looked up by city name or by coordinates with a radius. Each site includes GPU types, quantity, power capacity, chip generation, business category, and key customer, along with its location. By default only neo-cloud sites (GPU-as-a-service provid…
- [Health check](https://data.ornn.com/docs/api-reference/reference/health-check.md): Liveness and database check. Public, no key required.
- [List free-tier GPU models](https://data.ornn.com/docs/api-reference/reference/list-free-tier-gpu-models.md): Returns the GPU indices available on the public (no-key) tier. Pair this with `/api/gpu-types` — the full catalog Ornn tracks — to render the free-vs-paid split. The response does not vary by caller tier, so a keyed caller can still ask "what is free?" to build pricing or upsell UI.
- [List supported GPU models](https://data.ornn.com/docs/api-reference/reference/list-supported-gpu-models.md)
- [Get the latest token price index by lab](https://data.ornn.com/docs/api-reference/token-prices/get-the-latest-token-price-index-by-lab.md): The Ornn Token Price Index (OTPI): a daily, volume-weighted blended price per million tokens for each major model lab. Returns the latest settled day by default, a specific `date`, or a `startDate`/`endDate` range. Use `lab` to filter to a single lab. Without a key, the free tier covers 4 labs — `an…
- [List free-tier OTPI labs](https://data.ornn.com/docs/api-reference/token-prices/list-free-tier-otpi-labs.md): Returns the OTPI labs available on the public (no-key) tier as `{ lab, label }` pairs. Pair this with `/api/token-types` to render the free-vs-paid split. The response does not vary by caller tier.
- [List tracked OTPI labs](https://data.ornn.com/docs/api-reference/token-prices/list-tracked-otpi-labs.md): Returns every lab published by the Ornn Token Price Index (OTPI) as `{ lab, label }` pairs. Use this as the source of truth for valid `lab` values on `/api/otpi` and `/api/workload` instead of hard-coding them or discovering them from `400` errors.
- [Get the Workload Cost Index by provider](https://data.ornn.com/docs/api-reference/workload-cost/get-the-workload-cost-index-by-provider.md): The Workload Cost Index: daily settled cents per request, a mix-purged like-for-like index, and tokens per request for each closed-lab provider (`anthropic`, `openai`, `google`). With no `startDate` or `endDate` this returns the entire settled history for every provider and model family, so pass a r…
- [Authentication](https://data.ornn.com/docs/authentication.md): How to authenticate with the Ornn Data API using a Bearer API key, and which endpoints require one.
- [Chart prices in JavaScript](https://data.ornn.com/docs/chart-javascript.md): Fetch a GPU price series and render it in the browser.
- [Datacenter Sites](https://data.ornn.com/docs/datacenter-map.md): What the neocloud datacenter map shows: GPU-site locations, operators, capacity, and buildout timelines.
- [Errors & status codes](https://data.ornn.com/docs/errors.md): The error format, HTTP status codes, and common error responses for the Ornn Data API.
- [Introduction](https://data.ornn.com/docs/introduction.md): Documentation for the Ornn Data API, with read-only access to current prices, historical series, and market signals across GPU, memory, power, and AI markets.
- [LLM Coding Analytics](https://data.ornn.com/docs/llm-coding-index.md): Daily public-GitHub activity — PRs opened, merge rates, and commits — for Claude Code, Codex, Copilot, Cursor, and Devin.
- [Manage API keys](https://data.ornn.com/docs/manage-api-keys.md): Create, view, and delete your Ornn Data API keys from the dashboard.
- [Memory Price Index](https://data.ornn.com/docs/memory-index.md): What the memory prices represent, the tracked memory types, and how they update.
- [Pull price history into pandas](https://data.ornn.com/docs/pandas-dataframe.md): Load a GPU price series into a pandas DataFrame and plot it.
- [The Price Index](https://data.ornn.com/docs/price-index.md): What the GPU index value represents, its units, granularity, and how often it updates.
- [Quickstart](https://data.ornn.com/docs/quickstart.md): Get an API key and make your first request to the Ornn Data API in under two minutes.
- [Rate limits](https://data.ornn.com/docs/rate-limits.md): How rate limiting works on the Ornn Data API and how to avoid it.
- [Supported GPUs & memory types](https://data.ornn.com/docs/supported-types.md): The GPU models and memory types tracked by the Ornn Data API.
- [The Token Price Index](https://data.ornn.com/docs/token-price-index.md): What the OTPI is, how it's computed per lab, how to read the daily settled values, and every parameter the endpoint accepts.
- [The Workload Cost Index](https://data.ornn.com/docs/workload-cost-index.md): Daily realized dollars per request for the closed AI labs, plus a mix-purged like-for-like index and tokens-per-request intensity.

## OpenAPI Specs

- [openapi](https://data.ornn.com/docs/openapi.yaml)