parlant
Parlant is an open platform for customer-facing conversational agents. Its selection question is whether a team needs to manage changing business rules, tool use and human escalation within a continuous customer conversation. Evaluate it with mixed topics and follow-up requests rather than a single FAQ response. A read-only order lookup is a useful initial pilot before allowing operations that change business state.
On November 4, 2026, we checked the README, public architecture overview and package declarations. PyPI listed version 4.0.1 with Python 3.10 through 3.14 support. This is a documentation-based review: we did not connect a model provider, Dashboard or real order service, run multi-turn conversations or measure latency, cost or policy adherence. The linked guide labels its acceptance matrix as a proposed pilot.
The documented engine selects relevant policies, tools and approved wording as context changes, with reviews for actions and messages where needed. Separate those responsibilities during evaluation. Business APIs still need their own authorization, ownership checks and error handling; model-facing rules do not replace them. Human handoff also needs an operator who can actually receive context and act, rather than a transfer message alone.
Read the full review, verification scope and sources →Features
Why choose it
Trade-offs
Maintenance check
- Last commit
- 2026-11-10
- Commits, last 90 days
- 100+
- Releases, last 90 days
- 2
- Latest release
- v4.0.4 (2026-11-08)
- Open issues
- 13
- Recent issues with at least one comment
- 3/4
Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.
Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.
Most discussed open issues
- #683 [Docs] Warn users when NLP service does not offer constrained sampling7 comments · 2025-12-21
- #819 [Bug] <SDK accepts cross-journey transitions and later crashes journey projection with `KeyError`>4 comments · 2026-06-23
- #818 [Enhancement] Add Valkey as a VectorDatabase backend3 comments · 2026-06-22
- #782 [Enhancement] Extend the DI container2 comments · 2026-04-21
- #236 [Enhancement] Guideline Statistics1 comments · 2025-01-16
Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.
Who it fits
- Your service has explicit policies and custom business-system requirements.
- You can maintain conversation tests, tool authorization and human escalation.
- A fixed business workflow already meets the requirement.
- You expect a framework to supply staffing or guarantee rule adherence without testing.
Quick start
Use cases
Alternatives
Comments
No comments yet. Be the first!