Home / Directory / Data & labeling / Vector / retrieval infra / LanceDB
LanceDB
Multimodal lakehouse from San Francisco, built on the open-source Lance format, that stores vectors, images, video, and text in one table for search and model training.
AI labs and product teams with large image, video, or embedding datasets that need search, curation, and training on the same data.
Teams that want a small hosted vector index with a published price, or buyers who prefer a managed service with a free tier and a sign-up form.
Verdict
LanceDB is a data infrastructure company in San Francisco, California, founded in 2022 by Chang She, an original co-author of the pandas library, and Lei Xu, a core HDFS contributor who led machine learning infrastructure at Cruise. Its multimodal lakehouse, built on the open-source Lance columnar format, combines vector, full-text, and SQL search with training data curation, and customers include Midjourney and Runway. It matters because AI teams often copy the same data across a vector database, a feature store, and training storage, and LanceDB lets one table do all three.
Score Breakdown
How LanceDB scores in the categories that matter to its buyers.
Pricing
LanceDB does not list prices for its commercial product. Buyers contact sales.
The Lance format and LanceDB open-source packages are free to use. The commercial Multimodal Lakehouse is quoted by LanceDB's sales team.
The Field at a Glance
How LanceDB compares with other Vector / retrieval infra vendors we reviewed, by Overall Score and relative typical engagement cost.
LanceDB scores 7.8 in Vector / retrieval infra, under Pinecone (8.3), Chroma (8.1), Zilliz (8.0), and turbopuffer (7.9), and ahead of Weaviate (7.6). Relative cost lands upper-mid for this subcategory.
Use-case matrix
| Use case | Fit | Notes |
|---|---|---|
| Large multimodal training data | Strong | Built for petabyte-scale image, video, and text datasets. |
| High-traffic vector search | Strong | Midjourney chose it for high-traffic search. |
| Dataset versioning | Strong | Every write is versioned, with branch and roll back. |
| Hybrid search with SQL filters | Strong | Vector, full-text, and SQL in one query layer. |
| Small hosted index for a prototype | Mixed | The open-source library works; the commercial product is sales-led. |
| Published commercial pricing | Weak | Enterprise pricing requires a sales call. |
Who it’s for
Good fit
- Generative AI labs
- Teams training on image or video data
- Code and document search at large scale
Poor fit
- Prototypes that need a free hosted tier
- Buyers needing a public price list
- Teams with only a few thousand vectors
Review Excerpts
Below are excerpts from public reviews. Our team scoured public reviews, forums, and chat rooms to get a balanced view of customers' experience with this company. Paid reviews and pay-for-play sites such as Clutch were excluded.
“Lance transformed our model training pipeline at Runway. The ability to append columns without rewriting entire datasets, combined with fast random access and multimodal support, lets us iterate on AI models faster than ever. For a company building cutting-edge generative AI, that speed of iteration is everything.”
“Midjourney generates breathtaking imagery for millions of users worldwide. Vector search is critical infrastructure that allows us to better serve our users. We evaluated multiple solutions and LanceDB was the only one that could meet the high-traffic and large scale requirements we had. We couldn't be happier with our decision.”
LanceDB's pricing link routes to a contact form, so teams sizing the commercial lakehouse have to talk to sales before they see any cost.
Company Facts
- Founded
- 2022 (CB Insights)
- Headquarters
- San Francisco, California, United States (CB Insights)
- Employees
- About 35 (Y Combinator)
- Total funding raised
- $38.5M (CB Insights)
- Key investors
- Swift Ventures, Y Combinator, CRV, Zero Prime Ventures, Databricks Ventures (CB Insights)
- Accelerator
- Y Combinator, Winter 2022 batch (Y Combinator)
- Website
- lancedb.com
Key Leadership
Chang She
Co-founder and CEO
Methodology
This page is an independent evaluation of LanceDB for buyers comparing options in vector / retrieval infra. AI Industry Reviews accepts no sponsorships, advertising, or pay-for-placement fees. LanceDB did not pay for this review.
What we scored
The headline number is an Overall Score on a 0-10 scale. Eight criteria sit under it in two groups.
- Buyer outcomes (for vector / retrieval infra)
- Scale for multimodal data
- Search and training on one table
- Open-source foundation
- Pricing transparency
- Company & commercial
- Innovation & product leadership
- Project management & communication
- Pricing
- Contract fairness
Pricing measures whether the price looks fair for the value delivered, including packaging and renewal friction that show up in real buying cycles.
Score Composition
| Input | Weight | What it covers |
|---|---|---|
| Reviews | 40% | A proprietary read of what practitioners say about likes, complaints, and day-to-day use, including public review sites, forums, and private chat rooms. Paid reviews and pay-for-play sites such as Clutch are out of scope. |
| Product | 35% | Hands-on look at screens and workflows. |
| Pricing | 15% | Whether the price looks fair for what you get. |
| Docs & training | 10% | Docs, tutorials, and training material. |
How we balanced the evidence
The Overall Score is the simple average of the eight criteria. Recommendation language follows that score and the fit pattern described above.
Scope
LanceDB is graded here as vector / retrieval infra. Criteria scores can move as more review volume and product checks are added.