Skip to main content
Momentic uses several specialized AI agents. Each is versioned independently, so you can upgrade one without changing the others. Older versions stay available for backwards compatibility; set a specific version per agent in your momentic.config.yaml ai.agentConfig block. Momentic generally announces agent deprecations at least 90 days in advance.
When upgrading an agent version, we recommend testing the migration in a branch first. In rare cases, AI model decisions can vary from version to version.
The recommended agents use the latest SOTA models. They are specifically tuned to strongly respect memory (insights saved from past runs) as well as custom knowledge base entries. All agents have built-in fallback mechanisms to gracefully handle AI provider outages and disruptions. During such downtime, agents may incur additional latency or exhibit different behavior. However, Momentic uses multiple mechanisms such as memory and knowledge base, and failure recovery to reduce the frequency and impact of AI non-determinism on your tests.
AgentRecommendedp50 latencyOlder versions
locatorv43.6sv1, v2, v3
assertionv42.8sv1, v2, v3
visual-assertionv42.7sv1, v2, v3
text-extractionv34.2sv1, v2
failure-recoveryv2Variable (multiturn)v1

What each agent does

locator
Locates elements from a natural language description, powering Click, Type, and Element check steps. Strongly respects single quoted text values, builds more accurate caches, and achieves 15% lower latency compared to v3. Scores 10% better on our toughest web automation benchmark.
assertion
Evaluates natural language statements against a snapshot of the page, powering AI check steps. Strongly respects single quoted text values and discerns visual detail more accurately. Improves by 8% on our toughest web assertion benchmark with 22% reduced latency compared to v3.
visual-assertion
Evaluates natural language statements from a viewport screenshot. Excels at positional reasoning compared to previous versions. Scores 12% better on our toughest visual understanding benchmark.
text-extraction
Extracts structured data from the page given a JSON schema, powering AI extract steps. Handles nested objects and arrays.
failure-recovery
Generates and executes recovery steps when a recoverable failure is detected. Requires ai.failureRecovery.