Lochless vs warehouse AI functions

Data platforms let you run an LLM over every row to answer a yes/no question about your text. Lochless does the same job with an engine built for finding relevant text, at a fraction of the price.

Price per 1M tokens scanned

Lochless: $0.10, plus $0.005 per page indexed
Databricks ai_query: about $0.50
BigQuery AI.GENERATE_BOOL: about $0.86
Snowflake AI_FILTER: about $3.30

Our own test, October 2026. Snowflake and Databricks prices are our estimates from public list prices and measured token counts. BigQuery is measured tokens at list price. Your price depends on your plan, contract and region. The platforms have no indexing step.

Accuracy

Same 1,000 documents: 200 each from five public datasets (NQ, BoolQ, IMDB, SST-2 and Quora), the same ones Snowflake used to evaluate its AI_FILTER. Same yes/no question per dataset. Average F1: Lochless 0.875 on the live API, Snowflake 0.907, BigQuery 0.908, Databricks 0.905.

Lochless is about 3 points behind the platforms on average. The large difference is price. Small samples carry roughly plus or minus 0.01 to 0.03 F1 of noise.

When to pick which

Your text is already in the warehouse and volume is small: the built-in function is simplest.
You scan a lot of text, or your LLM or agent needs to query it directly: Lochless costs far less per token.
Neither handles questions that need outside world knowledge well. Lochless works only from what is in your text.

Cost control

Every query has a read limit, $1 by default. A free dry run returns the exact price first. Billing is prepaid and a $100 monthly spend cap is on by default.

Try it

Sign in with Google at console.lochless.io. Your workspace opens right away with $5 of free credit. No card needed. Then follow the quickstart.

No account yet? Point any MCP client at https://api.lochless.io/demo/mcp to try the same tools free on 20 public US founding documents.

Start free Read the quickstart