What Vespa.ai does

Vespa.ai AS is a Norwegian private company that operates a unified AI search platform combining vector and tensor search, lexical retrieval (BM25, WAND), and structured data queries with a multi-phase machine-learned ranking pipeline, executed against in-memory indexes for low-latency inference at scale. The technology originated inside Yahoo in 2011, where it still powers more than 150 applications at approximately 800,000 queries per second reaching around one billion users, and was spun out as an independent company in October 2023 with a $31 million Series A led by Blossom Capital. Vespa.ai is headquartered in Trondheim, Norway, with 51-100 employees, and serves enterprise customers across AdTech, Digital Commerce, FinTech, Market Intelligence, Health and Life Sciences, and Travel and Hospitality; named production deployments include Perplexity, Spotify, Yahoo, Farfetch, Vinted, Otto, Qwant, Kleinanzeigen, AlphaSense, Groupon, OkCupid, and Elicit.

Vespa's product portfolio is anchored by the open-source Vespa Engine on GitHub and delivered primarily through Vespa Cloud, a managed service that automates provisioning, autoscaling, canary deployments, and 24/7 monitoring, with a dedicated Vespa on AWS deployment available via AWS Marketplace and the AWS ISV Accelerate Program. Around the core engine, the company markets the Vespa Vector Database, Tensor Formalism, Vespa GenAI/RAG, Vespa Deep Research, Visual Retrieval (ColPali-based multimodal RAG), and the open-source RAG Blueprint template. The company generates revenue through usage-based Vespa Cloud subscriptions, enterprise on-premises contracts with support and SLAs for regulated industries, professional services and training, and an emerging channel via the March 2025 Vespa Partner Program.

Vespa.ai goes to market through a hybrid motion combining developer-led open-source adoption and a free Vespa Cloud trial with direct enterprise field sales, custom POCs, AWS co-sell, analyst-driven demand generation (GigaOm, BARC, ESG), and an in-person Vespa Live event series. The company's strategic focus is on real-time AI search, recommendations, RAG, and multi-hop research workloads where it claims differentiated results such as a 50 percent infrastructure reduction and 2.5x lower latency at Vinted after migrating from Elasticsearch.

Name

Vespa.ai

Legal name

Vespa.ai AS

Website

https://vespa.ai

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

51–100 employees

Short description

Vespa.ai operates a unified AI search platform that fuses vector, lexical, and structured retrieval with multi-phase ML ranking, delivered as open-source software and a managed cloud service to enterprise customers in AdTech, Digital Commerce, FinTech, and Market Intelligence.

Ownership category

akta.pro rank

Product category

AI Search Platform / Vector Database Software

NAICS

Software Publishers (5132), Custom Computer Programming Services (541511)

SIC

Services-Prepackaged Software (7372), Services-Computer Programming Services (7371)

akta.pro primary industry

Enterprise AI Data & Knowledge Platforms (Vector Databases, Knowledge Graphs) (HDAEANAH)

akta.pro secondary industry

Model Deployment, Serving & Inference Platforms (HDAAABAF)

Keywords

• Vector database platform
• Ai search engine
• Machine-learned ranking
• Retrieval-augmented generation
• Real-time inference

Headquarters

HQ city

Trondheim

HQ country

Norway

HQ region

Europe

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Vespa Cloud subscription: Managed cloud service billed on a predictable, usage-based model tied to cloud resources (compute, storage, queries). Customers pay recurring fees with consumption scaling up or down, supporting land-and-expand as workloads grow.
2. Usage-based cloud consumption: Resource consumption (queries, documents indexed, nodes, data transferred) is metered and billed, aligning revenue with customer workload volume.
3. Open-source self-hosted (free, unpaid): Vespa's core engine is open source and downloadable at no cost; this is not a direct revenue stream but a top-of-funnel that converts users into Vespa Cloud or enterprise contracts.
4. Enterprise / on-premises licenses and support: For customers running Vespa in their own environment (e.g., large banks, telcos, regulated industries), Vespa sells enterprise contracts that bundle support, SLAs, and advanced features.
5. Professional services and training: Vespa offers training programs and implementation support for large deployments, generating one-time services revenue alongside subscriptions.

Pricing tiers

Go-to-market motion6 records

Distribution channels7 records

Marketing channels11 records

Core offering

Vespa.ai develops and sells an AI Search Platform for building and operating large-scale applications that combine big data, vector search, machine-learned ranking, and real-time inference within a single unified engine. The platform is delivered as Vespa Cloud (managed service with usage-based billing) and via Vespa on AWS, with the open-source Vespa Engine available for self-hosted deployments, and includes specialized modules for retrieval-augmented generation, deep research agents, and visual document retrieval.

Product overview

Vespa.ai offers a single unified AI Search Platform anchored by the Vespa Engine (open source) and delivered primarily through Vespa Cloud (with a dedicated Vespa on AWS deployment available via AWS Marketplace). Around the core engine, the product portfolio includes the Vespa Vector Database, Tensor Formalism, Vespa GenAI/RAG, Vespa Deep Research, Visual Retrieval (Visual RAG), and the open-source RAG Blueprint template. These named modules plug into the same distributed query pipeline rather than being separate products: the Vector Database and Tensor Formalism provide the retrieval and ranking primitives, while GenAI/RAG, Deep Research, and Visual Retrieval are application-level capabilities built on those primitives, and the RAG Blueprint codifies the reference implementation used by customers like Perplexity.

Differentiator

Problem solved

Functional benefit

Brands

• The RAG Blueprint: A modular application template for designing, deploying, and testing production-grade RAG systems, built on the same core architecture that powers Perplexity.
• Vespa Cloud

Products and services

• Vespa AI Search Platform Unified AI search platform for developing and operating large-scale applications that combine big data, vector search, machine-learned ranking, and real-time inference; combines retrieval, ranking, and real-time decision-making within a single query pipeline for enterprise engineering teams.
• Vespa Cloud Fully managed production deployment of Vespa, automating provisioning, security, upgrades, autoscaling, canary deployments, and 24/7 monitoring. Includes the Tune-Up Program and next-business-day developer support for enterprise customers.
• Vespa on AWS Vespa Cloud deployment available via AWS Marketplace and the AWS ISV Accelerate Program, leveraging AWS Graviton processors and AWS-native infrastructure for customer-facing AI search and retrieval.
• Vespa Engine (Open Source) Open-source core of Vespa distributed on GitHub, providing the C++ content node engine, Java container clusters, and application package framework that powers Vespa Cloud and self-hosted deployments; freely downloadable.
• Vespa Vector Database Vector and tensor-native database engine inside Vespa supporting any number of vectors per document, HNSW-indexed approximate nearest neighbor search at billion-vector scale, real-time updates without rebuilds, and unified hybrid retrieval with text and structured filters.
• Vespa GenAI / RAG Retrieval-Augmented Generation offering combining hybrid text-vector search, machine-learned ranking, and LLM integration to deliver accurate, real-time AI retrieval for production customer-facing applications at scale.
• Vespa Deep Research RAG architecture for AI agents performing iterative, multi-hop research workflows at machine speed, unifying vector, text, and structured retrieval with on-cluster inference to handle hundreds of retrievals per agent session.
• The RAG Blueprint Open-source modular application template codifying best practices for production-grade RAG, covering searchable unit design, hybrid retrieval, phased ranking, ML-driven reranking, evaluation methodology, and query profiles for multiple use cases.
• Vespa Visual Retrieval (Visual RAG) Visual RAG solution that combines text and image queries via vision language models such as ColPali, enabling multimodal search across PDFs, charts, scanned documents, and product imagery for insurance, healthcare, e-commerce, and financial services use cases.

Quantifiable outcome

• 50% infrastructure reduction at Vinted
• +7 more outcomes

Named customers19 records

Segments6 records

Ideal customer profiles5 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration2 records

AI capability11 records

Feature7 records

Scale indicators9 records

Recent moves8 records

Expansion highlights6 records

Direct peers

• Elastic (Elasticsearch): The dominant open-source search and analytics engine that Vespa explicitly displaces in customer migrations (Vinted, Kleinanzeigen). Competes head-to-head in enterprise search, log analytics, and increasingly in vector/hybrid retrieval for AI applications.
• OpenSearch (Amazon Web Services): AWS-forked, open-source search and analytics suite that competes directly with Vespa's hybrid retrieval and vector capabilities, and is bundled into the AWS ecosystem that Vespa also relies on for distribution.
• Pinecone: A leading managed vector database purpose-built for AI retrieval and RAG. Competes with Vespa's vector database capabilities for AI-native workloads, particularly where simplicity and pure vector search are prioritized over hybrid retrieval.
• Weaviate: Open-source vector database with hybrid search and generative-search modules, frequently compared with Vespa in the AI/RAG category and competing for the same developer mindshare.
• Milvus / Zilliz: Open-source vector database (Milvus) and its commercial steward Zilliz, competing with Vespa's vector and hybrid retrieval capabilities for large-scale AI search and RAG deployments.
• Algolia: Managed search-as-a-service platform widely used in digital commerce, competing with Vespa in product discovery, catalog search, and personalized search for e-commerce customers like the ones Vinted, Otto, and Farfetch chose Vespa for.
• Qdrant: Rust-based open-source vector database with managed cloud offering, competing with Vespa in the vector and hybrid search segment for AI applications and RAG pipelines.

Emerging players

• Typesense: Open-source, typo-tolerant search engine often positioned as a simpler alternative to Elasticsearch for commerce and site search; competes with Vespa for developer-friendly hybrid and on-site search workloads.
• Marqo: Vector and multimodal search engine that, like Vespa, emphasizes end-to-end retrieval (including image and text) and is frequently benchmarked against Vespa in visual RAG and e-commerce product search scenarios.

Broad incumbents

• Coveo: Enterprise search, recommendations, and generative AI platform (Coveo Relevance Cloud) used by large enterprises for commerce, service, and workplace search; overlaps with Vespa's enterprise RAG and personalization use cases.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks5 records

Key highlights7 records

Customer concentration

• Xx.com/vespaengine
• GitHubgithub.com/vespa-engine/vespa
• YouTubeyoutube.com/channel/UCVXw_f6UHff8-V9FA1LMIiw
• LinkedInlinkedin.com/company/vespa-ai/posts/?feedView=all

Compliance2 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Funding overview

Funding rounds1 record

Investors1 record

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Vespa.ai