Jetty

Name: Jetty
Brand: Jetty

AI Native

Jetty is a managed platform that runs AI agents as versionable Markdown Runbooks, executing sandboxed workflows with programmatic evaluations and bounded retries while exposing an OpenAI-compatible API for integration.

Est 2025|Raised $1.5M|42 Ref Customers

LinkedIn 7K|Website|FTEs 5|HQ Montreal, Canada

01Competitive Landscape

| Aspect | Jetty (vendor site) | LangSmith (vendor site) | Langfuse (vendor site) | |--------|---------------------|-------------------------|-------------------------| | Starting Price | Free tier with limited monthly runs | Published developer-tier with per-seat and usage-based pricing | Free hobby tier with limits; published core-tier pricing available from the vendor | | Typical use case | Teams seeking runbooks-as-Markdown, managed sandbox execution, and built-in evaluation/iteration loops | Teams building LangChain/LangGraph applications needing tracing/evaluations and broader LangChain ecosystem tooling | Teams wanting tracing/evals with a self-host option; often chosen as an alternative to LangSmith | | Key Strength | Runbook as a portable spec; OpenAI-compatible run endpoint; trajectories with retry/evaluation loop | Strong ecosystem integration; mature observability and evaluations; integrated with LangChain tooling | Cost-accessible entry; self-host option; straightforward tracing/evals | | Key Weakness | Limited independent reviews; unclear public trust/security artifacts; run-quota billing can be surprising | Can become expensive at scale (per-seat cost + usage units) | Data access limits on lower tiers; advanced enterprise governance may require higher tiers | - Narrative analysis - Jetty vs LangSmith/LangGraph - Jetty’s vendor comparison frames the product as suited to judgment-shaped work where the agent definition is a Markdown runbook; LangGraph is positioned for graph-shaped orchestration. - Independent view - This framing is plausible: organizations commonly combine a workflow engine (Temporal or LangGraph) with a quality layer (evaluations/observability). Jetty attempts to bundle execution, evaluation, and iteration into a single managed platform. - If an organization already standardizes on LangSmith + LangGraph, Jetty must show additive value (sandbox execution, runbook spec, optimization loop) rather than acting as a duplicate dashboard. - Jetty vs Langfuse - Langfuse is often adopted for tracing and evaluations; Jetty’s differentiation lies in running workflows (sandboxed runbooks) rather than only observing them. - If the primary problem is lack of visibility into agent actions, tracing tools may suffice. If the primary problem is reliably re-running and gating quality, Jetty’s runbook and evaluation loop may align more directly. - Jetty vs Braintrust - Braintrust emphasizes evaluations and experimentation; Jetty emphasizes runbooks and an execution harness. Organizations can combine Braintrust-style evaluations with a separate execution runtime; Jetty aims to reduce integration effort by bundling both. - Adoption trends - Community signals (e.g., a Hacker News mention) indicate early adopter usage, but public data are insufficient to infer broad market share.

01Competitive Landscape

Aspect	Jetty (vendor site)	LangSmith (vendor site)	Langfuse (vendor site)
Starting Price	Free tier with limited monthly runs	Published developer-tier with per-seat and usage-based pricing	Free hobby tier with limits; published core-tier pricing available from the vendor
Typical use case	Teams seeking runbooks-as-Markdown, managed sandbox execution, and built-in evaluation/iteration loops	Teams building LangChain/LangGraph applications needing tracing/evaluations and broader LangChain ecosystem tooling	Teams wanting tracing/evals with a self-host option; often chosen as an alternative to LangSmith
Key Strength	Runbook as a portable spec; OpenAI-compatible run endpoint; trajectories with retry/evaluation loop	Strong ecosystem integration; mature observability and evaluations; integrated with LangChain tooling	Cost-accessible entry; self-host option; straightforward tracing/evals
Key Weakness	Limited independent reviews; unclear public trust/security artifacts; run-quota billing can be surprising	Can become expensive at scale (per-seat cost + usage units)	Data access limits on lower tiers; advanced enterprise governance may require higher tiers

Narrative analysis
- Jetty vs LangSmith/LangGraph
  - Jetty’s vendor comparison frames the product as suited to judgment-shaped work where the agent definition is a Markdown runbook; LangGraph is positioned for graph-shaped orchestration.
  - Independent view
    - This framing is plausible: organizations commonly combine a workflow engine (Temporal or LangGraph) with a quality layer (evaluations/observability). Jetty attempts to bundle execution, evaluation, and iteration into a single managed platform.
    - If an organization already standardizes on LangSmith + LangGraph, Jetty must show additive value (sandbox execution, runbook spec, optimization loop) rather than acting as a duplicate dashboard.
- Jetty vs Langfuse
  - Langfuse is often adopted for tracing and evaluations; Jetty’s differentiation lies in running workflows (sandboxed runbooks) rather than only observing them.
  - If the primary problem is lack of visibility into agent actions, tracing tools may suffice. If the primary problem is reliably re-running and gating quality, Jetty’s runbook and evaluation loop may align more directly.
- Jetty vs Braintrust
  - Braintrust emphasizes evaluations and experimentation; Jetty emphasizes runbooks and an execution harness. Organizations can combine Braintrust-style evaluations with a separate execution runtime; Jetty aims to reduce integration effort by bundling both.
- Adoption trends
  - Community signals (e.g., a Hacker News mention) indicate early adopter usage, but public data are insufficient to infer broad market share.

Aspect

Jetty (vendor site)

LangSmith (vendor site)

Langfuse (vendor site)

Starting Price

Free tier with limited monthly runs

Published developer-tier with per-seat and usage-based pricing

Free hobby tier with limits; published core-tier pricing available from the vendor

Typical use case

Teams seeking runbooks-as-Markdown, managed sandbox execution, and built-in evaluation/iteration loops

Teams building LangChain/LangGraph applications needing tracing/evaluations and broader LangChain ecosystem tooling

Teams wanting tracing/evals with a self-host option; often chosen as an alternative to LangSmith

Key Strength

Runbook as a portable spec; OpenAI-compatible run endpoint; trajectories with retry/evaluation loop

Strong ecosystem integration; mature observability and evaluations; integrated with LangChain tooling

Cost-accessible entry; self-host option; straightforward tracing/evals

Key Weakness

Limited independent reviews; unclear public trust/security artifacts; run-quota billing can be surprising

Can become expensive at scale (per-seat cost + usage units)

Data access limits on lower tiers; advanced enterprise governance may require higher tiers

Jetty

01Competitive Landscape

02Market Map

01Competitive Landscape

02Market Map