ibou opens its search index via API at €1 per 1,000 units

Abondance reports that ibou, a French search engine published by the Évreux-based company Babbar, is opening its index to third parties through an API. The ibou Platform product returns results from the engine, with four endpoints offering the same sources at different levels of detail, from a bare URL to the full page content. Each call costs between one and five units, and the reported rate is €1 per 1,000 units.

That is, for the moment, the whole of what is public. The announcement as carried by Abondance does not state which sources are indexed, how fresh the results are, what volume limits apply, what the contract terms look like, or how coverage compares with the major engines. Those gaps matter before anyone budgets against it, and they are the first things a team should establish.

A metered index, not a free feed

The pricing structure is the interesting part. Calls are billed in units that vary with the amount of data returned, which means a team can choose between cheap, thin responses and more expensive, richer ones. In practice this pushes the cost decision down to the query level: harvesting millions of URLs is a different exercise from fetching full pages for a shortlist of candidates.

For acquisition work, that logic maps onto a few familiar uses. URL-level results are enough to build keyword-to-page maps, spot which domains surface on a query set, or monitor how a topic's results evolve over time. Full-content responses are closer to a scraping substitute, useful for enrichment when a page cannot be fetched directly or when structure matters more than raw text. Which of these is actually cheaper than existing options depends on figures the announcement does not provide.

What to check before committing

The main risk with any new index is silent weakness. A small engine may cover mainstream queries adequately and thin out on niche or long-tail terms, which is exactly where SEO and content teams need reliable data. Coverage bias would not show up in a demo; it shows up when you compare result sets.

A reasonable first week looks like this:

• Run a fixed sample of 50 to 100 queries you already track elsewhere, from head terms to long-tail, and compare overlap with the engines you rely on.
• Measure returned units per call type to turn the published rate into a real cost per 1,000 queries and per 1,000 full pages.
• Read the terms on storage, redistribution and rate limits before any result enters a product or a client report.

Treat the first pass as a pricing and coverage test, not a migration. If the overlap is high and the unit costs hold, the API becomes a plausible second source for monitoring and enrichment. If coverage collapses outside mainstream queries, the cheap unit price is beside the point.

The broader signal is that search data is increasingly sold as a metered utility. Teams that already track cost per data call will be able to judge this offer in an afternoon; those that do not will keep comparing headline prices that say little about the bill.