Quanteta Data Lab launches LLM API Price Index covering 314 models

Quanteta Data Lab has published an LLM API Price Index that tracks official API list prices and specifications for 314 models from 50 providers.

What the index covers

The June 4, 2026 edition of Quanteta’s LLM API Price Index lists per-token API prices, context windows, and provider-published capability scores across a broad set of commercial language models, according to Quanteta’s index page.

Quanteta Data Lab’s methodology page says the dataset currently covers 314 models and 50 providers. The same methodology says the index aggregates official per-token API prices and model specifications, excludes free endpoints, and cross-checks values against OpenRouter and LiteLLM.

That scope makes the index a reference point for teams comparing the published costs of large language model APIs. It focuses on list prices rather than private discounts or negotiated enterprise contracts, based on the methodology’s description of using official per-token API prices.

How Quanteta says it builds the dataset

Quanteta Data Lab says its methodology relies on official provider pricing and specification information. The methodology page also states that free endpoints are excluded from the index, which means the dataset is aimed at paid API offerings rather than no-cost access tiers.

The methodology notes that Quanteta cross-checks OpenRouter and LiteLLM values. Those checks may help identify discrepancies, but the source material describes them as cross-checks rather than replacements for official provider information.

The index also includes provider-published capability scores, according to Quanteta’s index page. Because those scores are described as provider-published, readers should treat them as supplier-reported indicators rather than independent benchmark results.

Licensing and citation

Quanteta Data Lab’s citation guide identifies Quanteta Data Lab as the author of the Quanteta LLM API Price Index. The guide says the data is available under a CC BY 4.0 license, while noting that prices remain the property of the providers.

That licensing note is relevant for researchers, analysts, and procurement teams that may want to cite or reuse the dataset. The citation guide indicates that attribution to Quanteta Data Lab is required under the stated license.

Why it matters

API pricing for language models can be difficult to compare because providers publish costs in different formats and update product pages over time. Quanteta’s index addresses that comparison problem by placing per-token list prices, context windows, and provider-stated capability information in one reference.

The index does not, based on the provided methodology, claim to measure real customer spending, latency, reliability, or privately negotiated pricing. It is best read as a structured view of provider-published API prices and specifications at the time Quanteta lists them.

For buyers and builders, that distinction matters. Published per-token prices are useful for initial model selection and budget estimates, but actual deployment costs can also depend on prompt length, output length, caching behavior, traffic patterns, and contract terms. Those factors are outside the claims made in the provided Quanteta source excerpts.

Editorial artwork brief

Hero image prompt: A clean editorial illustration of many abstract AI model tiles arranged on a neutral grid, with small token-like dots and comparison bars suggesting pricing and context length, modern data-journalism style, no logos, no brand names, no readable text.

Hero image alt text: Abstract illustration of AI model comparison tiles with token dots and pricing bars.