Semantic browser queries and structured extraction for developer-controlled Playwright workflows.
Control frameworklive
AgentQL selects elements and extracts structured page data within an authored browser workflow. It is neither a full native desktop agent nor proof of a general goal-execution service; remote browser hosting and Enterprise deployment are separate commercial scopes.
Developers replacing brittle extraction selectors with semantic queries
Teams that can validate a structured extraction schema
LOOK ELSEWHERE WHEN
You need autonomous cross-application desktop execution
You need a documented security assurance or data-residency guarantee not established here
Fit recommendations are editorial interpretations of the documented architecture.
Documented capabilities
CAPABILITY
DOCUMENTED STATE
SOURCE
interaction — wrapper
The Python SDK wraps a Playwright page to expose semantic querying; the surrounding script still controls navigation and task order.
control — elements
query_elements and get_by_prompt locate interactive elements; query_data returns extracted values and is not an interactive element handle.
extraction — output
query_data returns a dictionary shaped by the supplied query, allowing the caller to validate and process requested fields.
configuration — modes
The page API documents fast and standard response modes with fast as default; mode selection is a speed/accuracy tradeoff, not a measured reliability guarantee.
configuration — wait
The page API exposes timeout, wait_for_network_idle and include_hidden controls; hidden-data handling and waits should be chosen for the authorized extraction scope.
pricing — billing
Starter and Professional separately meter API calls and remote-browser hours. Trial allowances are distinct from recurring plan quotas.
Environment, deployment & control
Environment
browser
Full
desktop
Unknown
mobile
Unknown
Deployment
cloud
Unknown
local
Unknown
self Hosted
Unknown
Extensibility
api
Unknown
sdk
Full
mcp
Unknown
bring Your Own Model
Unknown
Control method
screenshots
Unknown
mouse Keyboard
Unknown
dom Or Accessibility
Full
code Execution
Unknown
Unknown = insufficient public evidence; No evidence = an explicit negative record; N/A = does not apply to this layer.
Trust, oversight & limitations
Documented oversight
human Approval
Unknown
audit Logs
Unknown
sandboxing
Unknown
prompt Injection Defense
Unknown
credential Controls
Unknown
Know the boundary
No desktop, mobile, retention, human-approval or compliance guarantee is inferred.
Fast and standard modes are documented choices, not independent accuracy measurements.
The zero-dollar Starter subscription does not make overage calls or browser time free.
See Billing meters & conditions below for source dates, units and freshness. No unified calculator quote is available for this product; separately billed services are not a single total price.
Evidence, scope & unresolved questions
Stack classification: L2. Verified means the cited source supports the statement; it does not mean COMPUTER USE independently tested the product.
verified
status
AgentQL publishes a current Python browser-query SDK and self-service Starter/Professional API plans; Enterprise is a separately scoped offering.
Verified against official documentation; not independently executed or security-certified.
agentql ↗ · Verified Sep 29, 2026 · Recheck every 14 days
Billing meters & conditions
Each amount belongs to the named service, plan and unit. Subscription, runtime, steps and tokens cannot be ranked as a single cheapest price. Rates expire after 14 days; unresolved conflicts are excluded.
The owner announced Starter/Professional call allowances and a 300-call trial, while preserving prior free-tier terms for existing users. This historical announcement is not a substitute for current pricing.