Exa 3.0
Exa 3.0 is the latest release of Exa's web search API for AI applications, featuring a larger embedding model, generally available Dynamic Highlights, and a substantially expanded index covering ecommerce, local businesses, news, sports, weather, flights, events, transit, and podcasts. The release introduces Exa Lite, a lower-cost search type priced at $4 per 1,000 requests, alongside Exa Auto at $7 per 1,000 requests.
Key facts
What is Exa 3.0?
The fastest, most accurate web search API, giving AI agents the context they need at a lower cost per request.
Who is Exa 3.0 best for?
AI developers, coding agent and chatbot builders, inference providers, GTM and enterprise teams
Does Exa 3.0 have a free plan?
According to official sources, a free plan is available. Official pricing
What are its main limitations?
- Pricing is usage-based per 1,000 requests, so costs scale with query volume
- Zero data retention, HIPAA, SSO/SCIM and custom QPS are limited to the custom Enterprise plan
- The free Starter tier is capped at 10 Search QPS and 50 agent concurrency
Limitations are based on product materials or editorial synthesis; confirm on the official site.
When was this information verified?
October 20, 2026
Sources
- Official homepageofficial homepage · Oct 20, 2026
- Official pricingofficial pricing · Oct 20, 2026
Key features
- ✓Larger embedding model trained with more compute and data than any previous Exa model
- ✓Dynamic Highlights generally available, using 95% fewer tokens than full page text while improving answer quality
- ✓Exa Lite search type at $4 per 1,000 requests, running on the same index as Exa Auto
- ✓Expanded index tracking over 2 trillion URLs and serving 100 billion pages
- ✓Structured data coverage: 1.27B ecommerce listings with prices and availability, 211M+ local businesses with hours and ratings
- ✓Real-time news, sports, weather and market data, plus flights, local events, transit and podcasts
- ✓Agent API for asynchronous deep research runs with structured outputs and citations
- ✓Monitors API to identify fresh web events at a specified cadence and receive webhook updates
Use cases
- →Powering coding agents with low-latency search across repos, docs and changelogs
- →Grounding chatbots and voice agents with current web context
- →News monitoring and web change detection via scheduled searches and webhooks
- →Data enrichment and people/company search using structured index data
- →Model training and inference pipelines that need token-efficient page highlights
- →Cost-sensitive applications using Exa Lite at $4 per 1,000 requests
Pricing
Starter
$0
- · $10 credits every month plus a $10 onboarding bonus
- · No payment method required
- · MCP server access, Claude Connector, ChatGPT plugin
- · 50+ integrations and access to all endpoints
- · 10 Search QPS and 50 agent concurrency
- · Web, people, company and scholarly works indexes
Developer
Pay as you go (Search from $4/1k requests)
- · Everything in Starter
- · Usage-based billing with no commitment
- · Standard email support
- · Pay via card or bank account
- · SOC 2 Type II
- · Up to 25 Search QPS and 50 agent concurrency
- · Self-serve Exa Connect data providers
Enterprise
Custom
- · Everything in Developer
- · Custom MSA and DPA
- · Zero data retention
- · SOC 2 Type II and HIPAA
- · Monthly invoicing
- · SSO and SCIM for Exa Dashboard
- · Organization unified billing
- · Enterprise support with dedicated Slack channel and SLAs
- · Custom Search QPS and Agent concurrencies, up to 1,000 results per search
- · Custom tailored indexes and premium Exa Connect data providers
Free Starter plan with $10 monthly credits plus a $10 onboarding bonus; usage-based Search from $4 per 1,000 requests (Lite) or $7 per 1,000 (Auto), with custom Enterprise pricing.
Pros
- +Reports ranking first on every benchmark tested, spanning coding, finance, deep research and company intelligence
- +Dynamic Highlights included at no additional charge and uses 95% fewer tokens than full page text
- +Two search tiers (Lite at $4/1k, Auto at $7/1k) set a stated new Pareto frontier for cost vs. quality
- +Free Starter plan with $10 monthly credits plus a $10 onboarding bonus, no payment method required
- +Broad integration surface: MCP server, Claude Connector, ChatGPT plugin and 50+ integrations
Cons
- −Pricing is usage-based per 1,000 requests, so costs scale with query volume
- −Zero data retention, HIPAA, SSO/SCIM and custom QPS are limited to the custom Enterprise plan
- −The free Starter tier is capped at 10 Search QPS and 50 agent concurrency
Positioning
- Core value: The fastest, most accurate web search API, giving AI agents the context they need at a lower cost per request.
- Ideal for: AI developers, coding agent and chatbot builders, inference providers, GTM and enterprise teams
- Product type: API (Search, Contents, Agent, Monitors), MCP server, Claude Connector, ChatGPT plugin, 50+ integrations
Frequently asked questions
What is Exa 3.0?
Exa 3.0 is Exa's latest search release, which the company describes as its largest improvement to search quality to date. It includes a larger embedding model, generally available Dynamic Highlights, a significantly expanded index, and the new Exa Lite search type.
How much does Exa 3.0 cost?
Search pricing starts at $4 per 1,000 requests for Exa Lite and $7 per 1,000 requests for Exa Auto. Deep Search is $12 per 1,000 requests, Monitors is $15 per 1,000 requests, Contents is $1 per 1,000 pages per content type, and Answer is $5 per 1,000 requests. A free Starter plan includes $10 in credits each month plus a $10 onboarding bonus.
What is Exa Lite?
Exa Lite is a new search type priced at $4 per 1,000 requests, built for cost-sensitive applications. It runs on the same index as Exa Auto, which uses more compute across candidate generation, highlights and ranking for higher quality.
What are Dynamic Highlights?
Dynamic Highlights is a feature that selects tokens across all search results instead of extracting highlights from each page, giving more to the best pages and cutting repeated information. It is generally available at no additional charge and uses 95% fewer tokens than full page text.
How is Exa 3.0 delivered?
Exa 3.0 is available through the API, and the free Starter plan also includes MCP server access, a Claude Connector, a ChatGPT plugin and 50+ integrations.
You may also like
Curated suggestions based on similar categories.
RunInfra
RunInfra is a RightNow product offering hosted APIs for open models, built for AI agents. It uses session-aware routing to keep repeat calls on warm replicas, reporting high cache hit rates and lower cached-input pricing, with one API key compatible with the OpenAI and Anthropic SDKs.
AgentAccessDK
AgentAccess is an open-source, community-driven directory of machine-accessible interfaces (APIs, CLIs, MCPs, browser automation) for Danish digital systems, serving as the infrastructure layer for AI agents connecting to Denmark.
AISA
AIsa is a unified API gateway and transaction network for AI agents, providing access to 1000+ APIs, skills, and LLMs through a single key, and supporting autonomous micropayments via Circle Nanopayments and the Machine Payments Protocol (MPP).
HiAPI
HiAPI is a developer-first AI API platform that provides access to multiple leading image, video, and audio generation models through a single API key. It features a unified async task API, persistent artifact storage, callbacks, and transparent pay-as-you-go pricing with top-up packages.
Sinch Conversation API
Sinch Conversation API is a unified messaging API that connects businesses to 14 messaging channels (SMS, WhatsApp, RCS, Facebook Messenger, and more) through a single integration. It enables omnichannel customer communication with built-in AI for intent and sentiment analysis, automatic channel fallback, and global scale.
Removal.AI
AI-powered background removal tool that instantly removes backgrounds from images, offers editing tools, API integration, and credit-based pricing for high-resolution downloads.