Marqo
Ecommerce search and discovery trained on each retailer's own shopper data
Overview
Marqo is an ecommerce search and product-discovery platform that fine-tunes an embedding model on each retailer's catalogue and clickstream, then ranks search, category pages and recommendations for conversion. It began as an open-source vector search engine, now called Marqo Classic and deprecated since November 2025. Pricing is quoted after a demo.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
What you can produce with Marqo
- Ecommerce site search ranked for conversion by a model fine-tuned on the retailer's own clickstream and purchase data
- AI-ranked category and listing pages, collections and product recommendations
- A conversational shopping agent and visual search over the catalogue
- Merchandising rules (boosts, buries, pins and filters) layered over AI rankings
- Shopify, Adobe Commerce and Salesforce Commerce Cloud connectors plus a direct API
- Controlled A/B tests against the retailer's existing search provider
Marqo has turned from a general open-source vector search engine into a focused ecommerce search vendor: a pixel captures shopper behaviour, a per-retailer model is trained on it, and search, category pages, recommendations and a shopping agent are ranked for conversion, with Shopify, Adobe Commerce and Salesforce Commerce Cloud connectors and A/B tests against an incumbent.
The results it cites are its own case studies, and pricing only comes after a demo. The old open-source engine still exists under Apache 2.0 but is deprecated, so it is no longer a sensible free starting point. Worth testing for mid-market and enterprise retailers; not for developers who want a vector database.
What we found
Marqo is no longer the product this page first described. It launched in 2022 as an open-source vector search engine from Melbourne, built by former Amazon engineers Tom Hamer and Jesse Clark, that generated embeddings, stored them and searched them through one API. That engine now lives on as Marqo Classic. Its GitHub README has carried a deprecation notice since November 2025, and Marqo's documentation says Classic is no longer actively developed. The code remains under the Apache 2.0 licence; the last tagged release, 2.26.0, came out in April 2026.
The current Marqo is an ecommerce search and product-discovery platform for mid-market and enterprise retailers. A tracking pixel collects clicks, add-to-carts and purchases, and Marqo fine-tunes an embedding model for each retailer on that data, then uses it to rank search results, category pages, recommendations and a conversational shopping agent. Connectors exist for Shopify, Adobe Commerce and Salesforce Commerce Cloud, alongside a direct API, and the documentation pitches Marqo as a replacement for Solr or Elasticsearch builds, Searchspring and Klevu.
Evidence of results comes from Marqo itself. Its homepage quotes case-study uplifts of roughly 10% to 28% in search conversion and revenue, including a 17.7% conversion uplift at sneaker marketplace KICKS CREW, and shows Mejuri, Kogan, SwimOutlet, Poshmark, Redbubble and Shutterstock among its customers. These are vendor-reported A/B results, not audited ones.
Pricing is not published: the pricing link leads to a demo form, and the AWS Marketplace listing sells contracts of one to 36 months priced per storage shard. The last announced round was a $12.5 million Series A led by Lightspeed in February 2024, taking total funding to $17.8 million, when the headquarters moved to San Francisco. The founders have since also set up Celeris, a separate lab building fast language models. For a retailer replacing site search, Marqo is worth a test; developers wanting a vector database should look elsewhere.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Marqo's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Marqo unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Marqo directly →
Spotted something out of date? Suggest an update →
More in Research & Data
