For developers
OramaJS
Free · open source
Add fast, relevant search to your own application with a free, open-source library.
10.5kGitHub stars
Search and AI infrastructure
Search that answers your customers and your team
Orama makes the content your customers read and the data your team runs on searchable in one place, and turns a question into a straight answer with the sources behind it.
OramaJS on GitHub · 10.5k starsorama · search preview
what did we promise northwind on support response times?
AI answer
The signed MSA promises a one-hour first response for priority-one issues and four hours for everything else. The 2025 renewal kept both targets and added weekend coverage.
grounded innorthwind msasupport sla policyrenewal notes · march 2025
- Northwind · master service agreement
- Support response times by plan tier
- Renewal call notes · March 2025
10.5k
GitHub stars on OramaJS, our open-source search library
Orama Search
Built by the team behind Orama Search
Open source
Read the core library's code before you spend anything
Your cloud, on-prem, or dedicated
Enterprise deployments run where you choose
The Orama suite
Start free. Move up when you need more.
Every Orama product does one job well, from an open-source library to an enterprise platform. Pick the step that matches where you are. The next one is there when you need it.
For business teams without a search team
Orama Cloud Pro
Flat monthly price
Connect your content and get production search with AI answers, with no infrastructure to run.
For enterprise platform and AI teams
Orama Zetta
Contact sales
The enterprise platform: search, answers, and AI models trained on your own data, run in your cloud, on-prem, or dedicated.
See it work
From a search box to an answered question
Ordinary search returns a list and leaves the reader to dig. Agentic search plans its own retrieval across your content, then returns an answer with the sources behind it. For the person asking, that is the difference between a list to dig through and an answer to act on.
Agentic: plans retrieval across your sources and returns an answer with citations.
Your contentSource
embedEmbedderModel
writeSearch indexStore
Grounded in your content
QuestionSource
Answer agentAgent
retrieveSearch indexStore
answerCited answerOutput
A customer who gets an answer keeps shopping instead of filing a ticket. A teammate who gets an answer stops asking around and gets back to work. That is the business case for better search.
Beyond search
Where teams take it next
Search is the front door. Once your content and signals live in one system, other work gets easier. Where something is still early, the label says so.
Early accessOrama Zetta
Churn prediction
See which accounts are likely to leave while there is still time to act, with the drivers behind each score. Models train and validate on your own data before anyone relies on them.
Talk to sales: Churn predictionEarly accessOrama Cloud Pro
Automated outbound
Outreach drafted from your own content and account context, grounded in retrieved sources, and reviewed by a person before anything is sent.
Start Pro: Automated outboundIn developmentOrama Zetta
Agentic RL models
Reinforcement-learning agents trained and governed on your data, with explicit approval steps before deployment. No availability date yet, and no performance claims.
Talk to sales: Agentic RL modelsWhy Orama
Built by search people
Search is where we started
Orama comes from the team behind Orama Search and the open-source OramaJS library. We build search first and add AI where it earns its place.
An index that keeps learning
The adaptive database learns from the data it holds and the queries it serves, so relevance improves over time instead of staying static.
You choose where it runs
Take the managed cloud when you want nothing to host. For enterprise deployments, Orama Zetta runs in your cloud, on-prem, or dedicated.
AI you can check
Answers cite their sources. Models are validated before deployment, and a person owns the decision to act on what they say.
Orama Zetta · enterprise platform
When search becomes infrastructure
Orama Zetta is the enterprise platform: an adaptive database with auto-AI, run inside your own environment. In business terms, your data and its patterns live in one system that searches it, answers from it, and trains models on it, with validation before anything reaches production.
- Deploy in your own cloud, on-premises, or on dedicated infrastructure
- One system for search, answers, and models on your own data
- Models proposed, trained, and validated before deployment
- Sales-led, scoped to your deployment
No public pricing. A call scopes the deployment first.
Your cloud, on-prem, or dedicated
Auto-AIModel
modelsAgentsAgent
trainsAdaptive databaseStore
datagoverns outputValidation & controlsControl
Pricing
Ways to start
The library is free and open source. The managed cloud is a flat monthly price plus one-time onboarding, with annual prepay available. Enterprise is scoped on a call, because deployments differ. Exact prices come at checkout or on the call.
Orama Cloud Pro
Flat monthly price
Managed search with AI answers, plus one-time onboarding with our team. Annual prepay available. The exact price shows at checkout, or on a call.
Orama Zetta
Contact sales
The enterprise platform, deployed in your cloud, on-prem, or dedicated.
Ready when you are
Start on Orama Cloud Pro today, or book a call and we will scope it with you.