Restb.ai
Computer-vision API that reads real estate photos at MLS scale
Overview
Restb.ai runs computer vision on property photos — tagging rooms, scoring condition and quality, flagging watermarks or duplicates, drafting listing descriptions — sold as an API to MLSs, appraisal platforms and data providers such as Cotality (ex-CoreLogic) and ICE, not to individual agents. Clear Capital acquired it in May 2026 and kept the brand.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 30 Sep 2026 and may be out of date or incomplete.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 Restb.ai
- Room-type and feature tagging (image classification)
- Property condition and quality scoring
- Duplicate and watermark detection
- Auto-generated listing descriptions
- Visual similarity / comparable-property matching
- MLS photo-compliance flagging
- RESTful API returning JSON, with batch processing and free proof-of-concept trials
- UAD 3.6-aligned condition and quality outputs (they do not generate the appraisal itself)
Restb.ai is infrastructure, not an app — a computer-vision API that MLSs, AVM providers and appraisal-tech platforms plug into rather than something an agent logs into. Its niche is specific enough that general-purpose vision models don't compete well: standardized room classification, condition and quality scoring calibrated for appraisal use, and watermark and duplicate detection at listing-photo volume — the company says more than 2 billion photos a month. Clear Capital's acquisition in May 2026 gives it a larger appraisal-technology parent while keeping the brand, and integrations with Cotality and ICE lend real credibility. Treat its accuracy claims as vendor-reported: the widely quoted 98% figure comes from Fannie Mae's own review, which does not name Restb.ai, and its white paper finding a third of appraisals at high adjustment risk makes the case for its own product. Pricing is sales-gated — a monthly fee with an API-call allowance and overage — with free proofs of concept. This is a developer/integrator tool, not an end-user product.
What we found
Restb.ai has a new owner. On 12 May 2026 Clear Capital, a US property-valuation and appraisal technology company, announced it had acquired Restb.ai, keeping the brand. Restb.ai now says it processes more than 2 billion property photos a month and reaches more than a million agents through 26 MLSs. Like most computer-vision APIs sold to MLSs and appraisal platforms, it has no G2 or Capterra listing, so what can be checked publicly comes from partners and its own research.
The partner list is the strongest signal of real-world traction: Cotality (formerly CoreLogic) since 2019, ICE (which bought Black Knight), Blackstone, Lemonade, Idealista and Stewart Valuation Intelligence are named as customers or integrators, and 2026 brought integrations with ApprAIz, BoxLi and Realsy.
Accuracy claims need care. A figure often attached to Restb.ai — 98% accuracy in spotting appraisal defects — describes Fannie Mae's own image-recognition review; Restb.ai quotes Fannie Mae's Loan Quality Center as finding the model prediction 98% accurate, but Fannie Mae's release does not name the vendor. A 99% search-accuracy figure dates from a 2018 company blog. Treat both as vendor-reported.
The company's May 2025 white paper is more revealing: analysing 1,271 appraisals and 6,495 comparables with its own computer vision, it found a third of appraisals carried a high condition- or quality-adjustment risk and nearly three-quarters a medium one. That is Restb.ai quantifying the problem it sells a fix for — useful, but vendor research rather than independent validation. Deployment figures in a HousingWire Tech100 profile — a 50% cut in manual review, a 50% drop in revisions and a 28% sale-price lift through Revive's platform — are also vendor-supplied.
Pricing is not published: the FAQ describes a fixed monthly fee with a set number of API calls and a per-call overage, batch pricing on request, and free proof-of-concept trials. Its UAD 3.6 tools help appraisers align with the new standard, though the company notes they do not generate the appraisal itself.
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 Restb.ai's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Restb.ai 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 Restb.ai directly →
Spotted something out of date? Suggest an update →
Alternatives to Restb.ai
More in Research & Data
