Nimble
by Nimble Way (The Data Company Technologies Inc.)
Expert-level web search agents that learn your domain, cutting AI research token spend roughly in half
Nimble is a web data and search platform built for AI agents rather than humans. Its Web Search Agents, launched July 29, 2026, autonomously learn a customer's domain - compliance auditing, lead generation, market or financial research - and return curated structured tables instead of raw pages, which is where the token savings come from.
Nimble is a real-time web search and data platform for enterprise AI systems, built by Tel Aviv-headquartered Nimble Way with offices in New York and London. Rather than handing an LLM raw HTML to parse, the platform uses AI-driven agents to browse sites, extract information and convert dynamic web content into structured datasets. On July 29, 2026 it launched Web Search Agents, its flagship product: domain-specialized research agents that autonomously learn a customer's use case and execute complex multi-step web research, returning curated data tables. Nimble reports the agents deliver a 21-point increase in answer quality and 51 percent fewer tokens spent per query against leading alternatives, and maintains auditable Search Plans showing exactly what was retrieved and why. The platform is reachable three ways - REST API, SDK and MCP - so agent frameworks can call it directly, and it exposes discrete Search, Extract, Crawl, Map, media-retrieval and Agent endpoints alongside managed data-service plans and pre-built structured feeds. Nimble reports powering more than 90 million searches daily and over 1.5 million developer users, with named enterprise customers including Databricks, Microsoft, Uber, Coca-Cola, Tripadvisor, L'Oréal, Semrush, Deloitte and LG AI Research. It is SOC 2 Type II certified, GDPR and CCPA compliant, and offers a zero data retention policy with PII masking, audit logging and encryption in transit. In February 2026 the company raised a $47 million Series B led by Norwest Venture Partners with Databricks Ventures participating, bringing total funding to $75 million.
Data and AI platform teams whose agents already depend on live web data and who are paying for it twice - once in scraper maintenance and again in tokens burned parsing raw pages.
Structured, domain-tuned research results with auditable Search Plans, at roughly half the tokens per query of feeding raw pages to an LLM.
At a Glance
- Category
- Data & Analytics
- Pricing
- Usage-based, Subscription
- Target Market
- CTOs, Chief Data Officers, Data Engineers, AI/ML Engineers, Enterprise Developers
- Deployment
- Cloud-only, API-based
- Headquarters
- Tel Aviv, Israel
- Team Size
- 51-200
- Customers
- Over 1.5 million developer users and more than 90 million searches daily; named enterprise customers include Databricks, Microsoft, Uber, Coca-Cola, Tripadvisor, L'Oréal, Semrush, Deloitte and LG AI Research
Key Features
- ✓Web Search Agents
Domain-specialized agents that self-learn a customer's use case and return curated data tables rather than raw search results.
- ✓Auditable Search Plans
Every run exposes exactly what was retrieved and why, giving regulated teams provenance for AI-sourced web evidence.
- ✓Search, Extract, Crawl and Map APIs
Discrete endpoints for querying, structured capture, full-site indexing and media retrieval, each metered separately.
- ✓MCP and SDK access
Agent frameworks call Nimble directly over MCP or SDK, so live web intelligence drops into an existing agent stack.
- ✓Token-efficient structured output
Returning parsed tables instead of raw HTML removes the LLM parsing step Nimble says accounts for about half of query tokens.
- ✓Enterprise data governance
SOC 2 Type II, GDPR and CCPA compliance, zero data retention, flexible PII masking, audit logging and encryption in transit.
- ✓Managed data services with warehouse delivery
Tiered plans provide concurrent agents, page credits, retention windows and native delivery into a data warehouse.
Capabilities
Use Cases
- •Autonomous market and competitive research
Agents learn the analyst's domain, scour relevant sources continuously and assemble curated comparison tables without hand-written scraping scripts.
- •Compliance and regulatory monitoring
Auditable Search Plans record exactly which sources were retrieved and why, which matters when an AI finding must be defended.
- •Lead generation and account enrichment
Structured extraction across company sites and directories builds and refreshes prospect datasets that would otherwise be maintained manually.
- •Grounding enterprise agents in live web data
MCP and SDK access lets an existing agent stack pull fresh public web intelligence at roughly half the tokens per query.
- •Building structured datasets for model training and analytics
Crawl and Extract endpoints turn dynamic web content into warehouse-ready tables for downstream analytics or fine-tuning pipelines.
Ideal For
Best For
- ✓AI agent platforms that need live public web data through MCP without each team building and maintaining its own scrapers
- ✓Compliance auditing, market analysis, lead generation and financial research workflows where curated tables beat raw page dumps
- ✓Enterprises where token spend on web research is a material and growing line item in the AI budget
- ✓Regulated teams needing auditable retrieval provenance, PII masking and zero data retention on public-web collection
- ✓Data engineering groups already on Databricks or Microsoft that want native warehouse delivery of web-sourced datasets
Not Ideal For
- ✗Individual developers and one-off projects - the cheapest managed plan is $2,500 per month billed annually, and the whole tier structure is built around concurrent agents and million-page credit blocks
- ✗Teams needing internal enterprise search over private documents; Nimble collects publicly accessible web data, not intranet content
- ✗Buyers who want turnkey simplicity, since the platform exposes six separate metered endpoints plus managed ETL rather than one interface
- ✗Organizations that require independently verified performance claims, since the 21-point quality gain and 51 percent token reduction are Nimble's own benchmarks
- ✗Simple, low-volume scraping needs that a cheap proxy provider or an open-source crawler already covers
Integrations
Deployment
Market & Ratings
Over 1.5 million developer users and more than 90 million searches daily; named enterprise customers include Databricks, Microsoft, Uber, Coca-Cola, Tripadvisor, L'Oréal, Semrush, Deloitte and LG AI Research
Market Analysis
Pros
- ✓Named production customers are unusually strong for the category - Databricks, Microsoft, Uber, Coca-Cola, Tripadvisor, L'Oréal and LG AI Research, with Databricks Ventures also an investor
- ✓Published, itemised pricing on both the API and managed-service tracks, which most competitors in this space do not offer
- ✓Genuine enterprise governance posture: SOC 2 Type II, GDPR and CCPA compliance, zero data retention, PII masking and audit logging
- ✓Reachable over MCP as well as REST and SDK, so it drops into an existing agent stack without glue code
- ✓$75M raised and reported scale of 90 million searches daily suggest the infrastructure is production-proven rather than a launch demo
Cons
- ✗The headline performance claims - a 21-point answer-quality gain and 51 percent fewer tokens per query - come from Nimble's own benchmarks and, as TechTarget notes, have not been independently verified
- ✗Analyst reception was measured rather than enthusiastic: Donald Farmer of TreeHive Strategy called the work solid but not particularly innovative against competitors, and Constellation's Michael Ni tied differentiation to whether Nimble actually delivers greater completeness and lower maintenance cost than traditional scraping platforms
- ✗Priced for medium-to-large organizations - the cheapest managed plan is $2,500 per month billed annually, and useful page-credit volumes sit at the $7,000 and $15,000 tiers
- ✗Data retention is a tier gate rather than a setting, capped at 7 days on Startup and 30 on Scale, so longer lookbacks force an upgrade
- ✗A crowded field with well-established incumbents - TechTarget names Apify, Bright Data, Grepsr, MixRank and Oxylabs competing on the same web-data collection problem
- ✗No verified aggregate user rating could be established: G2's Nimble review page blocks unattended requests, and no rating was recorded rather than repeating one that could not be opened
Pricing
Pay-as-you-go (API)
From $1.00/1,000 URLs
- ✓5,000 free web pages to start
- ✓Search API $5 per 1,000 search inputs (up to 100 results per request)
- ✓Extract / Crawl / Map $1.00-$1.45 per 1,000 URLs
- ✓Extract Template API $3.00 per 1,000 pages
- ✓Agent API priced by selected effort level
Startup (Data Services)
From $2,500/mo
- ✓5 concurrent agents
- ✓350K monthly web page credits
- ✓7-day data storage
- ✓Custom agent ETL and MCP integration
Scale (Data Services)
From $7,000/mo
- ✓10 concurrent agents
- ✓1.2M monthly web page credits
- ✓30-day data storage
- ✓Localized agents support
Professional (Data Services)
From $15,000/mo
- ✓20 concurrent agents
- ✓3M monthly web page credits
- ✓90-day data storage
- ✓Custom workflows
Enterprise
Contact for pricing
- ✓Unlimited concurrent agents
- ✓Custom data storage
- ✓Advanced security and SLAs
Two published tracks. Pay-as-you-go meters per unit of work - $5 per 1,000 Search API inputs, $1.00 to $1.45 per 1,000 URLs for Extract, Crawl and Map depending on driver complexity, and $3.00 per 1,000 pages for the Extract Template API - with the Agent API priced by the effort level you select and 5,000 free pages to trial it. Managed Data Services are annual-billed subscriptions gated on concurrent agents and monthly page credits: Startup $2,500/mo for 5 agents and 350K credits, Scale $7,000/mo for 10 agents and 1.2M credits, Professional $15,000/mo for 20 agents and 3M credits, Enterprise custom. Data retention length is itself a tier gate, running 7, 30 and 90 days respectively.
Security & Compliance
Connect
Sources
This page was written from 4 sources, 2 on domains other than nimbleway.com.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Quantum Metric Felix Agentic
Agents that watch your digital funnel, find what broke and price the damage
Zilliz Vector Lakebase
Unified vector lakebase merging real-time vector search, analytics and data lake queries on one copy of data
LanceDB
The multimodal lakehouse for AI — vector, full-text, and hybrid search at billion-row scale
Akeneo Agentic Ziggy
Agentic AI layer inside the Akeneo Product Cloud that enriches, governs, and orchestrates product data at catalog scale