← Directory

Parlant

Parlant is an open-source conversational control layer for building and maintaining customer-facing AI agents.

  • AI
  • Automation
  • Development
Parlant is an open-source conversational control layer for building and maintaining customer-facing AI agents. It gives developers a structured harness to write agent behavior as code, iterate on responses through conversation, and keep them consistent, compliant, and on-brand as complexity grows. It targets enterprise B2C and sensitive B2B interactions, with extensive traceability and explainability for troubleshooting production agents.

Platform availability

Available on 1 of 9 platforms

  • Unraid (not listed)
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

70/100 Good

Maintained
74last commit 104 days ago
Popular
8518,299 GitHub stars
Easy to install
251 install platform, install notes
Light to run
No data
Security check
88no serious findings, 2 minor, CI, tests

Based on 4 of 5 factors.

GitHub stars
18,299
Open issues
40
Last commit
2026-07-10
Latest release
v3.3.2
License
Apache-2.0
Activity
stale

Checked 3 days ago - source: GitHub (emcie-co/parlant)

Resources & Compatibility

Requirements not published yet.

First-install notes

  • Install the Parlant client library For Python, run pip install parlant-client; for TypeScript, run npm install parlant-client. This installs the client library your code uses to talk to Parlant, so without it your app can't create or control agents. source
  • Set Ollama development variables If you use Ollama (a local AI model runner), set the OLLAMA_MODEL, OLLAMA_EMBEDDING_MODEL, and OLLAMA_API_TIMEOUT environment variables (named settings your system reads). These choose the models and wait time, reducing missing-model or timeout errors. source
  • Check the test playground After starting the example server, open http://localhost:8800 to see the test playground. If it doesn't load, check that the server is still running and that port 8800 (a numbered local address) is not blocked. source

Fetched 3 days ago - each note links to its source

Alternatives (8)

Checked 3 days ago - source: AlternativeTo

Common questions

Can I switch NLP services while keeping existing Parlant config and data?

In 1.6 this was unsupported, but it was fixed in 1.7.0. Use the latest release.

Do I need a separate Parlant client library to call the server?

Yes. Use pip install parlant-client for Python or npm install parlant-client for TypeScript; without it your app cannot create or control agents.

Which environment variables should I set when using Ollama with Parlant?

Set OLLAMA_MODEL, OLLAMA_EMBEDDING_MODEL, and OLLAMA_API_TIMEOUT. These choose the models and wait time.

Which local port is used for the example server playground?

Use the test playground on port 8800 on localhost. If it does not load, check that the server is running and that port is not blocked.

What does production deployment of Parlant involve?

Production deployment requires changes beyond local setup. The guide covers Kubernetes with authentication, persistence, and scaling.

Answers sourced from github.com, github.com, github.com, github.com, www.parlant.io

Community

Collected 3 days ago - public community threads