Agentic WeeklyGet the weekly
Open Home Foundationv1.8.0Tested (smoke test)

Piper: what’s new in v1.8.0 and our verdict

Fast local neural text-to-speech that runs on CPU.

Official site

Release Ruler for Piper: marked at its Tested release date on the timeline of every AI model we track.

At a glance

Evidence
Tested (smoke test)
Released
Sep 4, 2026
Last verified
Oct 25, 2026
Pricing
Partly verified
Best for
AI voice and speech

Evidence and ranking

Evidence behind this verdict

  1. Sources read1 cited source, latest fetched Oct 25, 2026. (done)
  2. Facts re-checkedLast verified Oct 25, 2026. (done)
  3. Run by usSmoke test on Oct 25, 2026 (scope below). (done)

Passed Test log result: Passed. Read the log

Where we rank it (editorial)

Listed here but not yet ranked. Ranks are editorial picks, not benchmark scores.

Pricing: Open source; no pricing page read. (checked 2026-10-25)

Our verdict

Fast, private speech synthesis on a CPU; mind the GPL-3.0 license if you embed it.

Good for

  • Local voice output without a GPU
  • Home and offline projects

Skip if

  • You need very natural or expressive voices

Tested. Tested 2026-10-25 (America/Chicago). Piper 1.8.0 synthesised 10 sentences (about 41 s of 22.05 kHz audio) on CPU in 1.4 to 1.5 s each including process start; 3.5 s uv install, 167 MB. Test environment: Linux x86_64 virtual machine (Debian 13), 8 vCPU, 15 GiB RAM, no GPU; no vendor account, login or API key. Not tested: Voice quality as judged by a person (we only measured machine round-trip accuracy), other voices and languages, streaming, and GPU use; One voice only; a smoke test. These are smoke tests, not a benchmark. Full log: /tests/piper.

Evidence
Tested (smoke test, 2026-10-25, no LLM involved). A smoke test is not a benchmark.
What we ran
Piper 1.8.0 synthesised 10 sentences (about 41 s of 22.05 kHz audio) on CPU in 1.4 to 1.5 s each including process start; 3.5 s uv install, 167 MB.
Model used
none (no language model was involved)
Environment
Linux x86_64 virtual machine (Debian 13), 8 vCPU, 15 GiB RAM, no GPU; no vendor account, login or API key
Not tested
Voice quality as judged by a person (we only measured machine round-trip accuracy), other voices and languages, streaming, and GPU use; One voice only; a smoke test.
Date
(America/Chicago)

Best for

Key facts

Piper is a local text-to-speech engine. The repository piper1-gpl is GPL-3.0 licensed; v1.8.0 was tagged on 2026-09-04.

  • License: GPL-3.0 (per GitHub)
  • Platforms: Python; Linux, macOS, Windows
  • Pricing model: Open source
  • Free tier: Open source, free to use
  • Paid plans: Hosted or enterprise offerings not read
  • GitHub: OHF-Voice/piper1-gpl: 6,017 stars on 2026-10-25, last push 2026-10-25
  • Latest release we saw: v1.8.0, 2026-09-04
  • How to start: uv pip install piper-tts and download a voice.
  • v1.8.0 (2026-09-04): the newest release on the repository when we read it on 2026-10-25.

How to set it up: Get started with Piper

Install steps are exactly what we ran on 2026-10-25; the test log lists what was and was not covered.

  1. Install

    Package name on PyPI is piper-tts.

    uv pip install piper-tts

  2. Download a voice and speak

    Voices are separate downloads (about 60 MB for the one we used).

    python -m piper.download_voices en_US-lessac-medium && python -m piper -m en_US-lessac-medium -f out.wav -- "Hello"

Pros and cons

Pros

  • Fast on CPU: about 1.4 s per sentence in our run
  • Offline

Cons

  • GPL-3.0 license
  • Voices are separate downloads
  • We judged accuracy through transcription, not how natural it sounds

Alternatives and comparisons

  • faster-whisper: Speech-to-text library that runs OpenAI Whisper models with CTranslate2.
  • Ollama: The common way to run open models locally on macOS, Windows and Linux, with a cloud option and an MIT-licensed open-source core.
  • AssemblyAI: Speech AI APIs for transcription and understanding; start free, pay as you go.
  • Cartesia: Realtime voice AI: text-to-speech, speech-to-text in 44 languages and voice agents, with a free plan and credit tiers.

FAQ

What is Piper?

A local neural text-to-speech engine that runs on CPU. The piper1-gpl repository is GPL-3.0 licensed.

How fast is Piper on CPU?

About 1.4 to 1.5 s per sentence including process start for 3 to 5 s of audio in our test. See /tests/piper.

What we have not verified

  • Other voices and languages
  • Perceived naturalness

Sources

Last verified · Agentic Weekly

  • AssemblyAI: Speech AI APIs for transcription and understanding; start free, pay as you go.
  • Cartesia: Realtime voice AI: text-to-speech, speech-to-text in 44 languages and voice agents, with a free plan and credit tiers.
  • Deepgram: Voice AI APIs for speech-to-text, text-to-speech and agents with $200 in free credit.

How to cite this page

Agentic Weekly, "Piper v1.8.0: what's new, verdict & setup", last updated 2026-10-25, https://theagenticweekly.com/tools/piper

Keep the Reported or Tested label when you quote a verdict. Plain-text version: https://theagenticweekly.com/tools/piper.md. More for AI assistants: /ai, /tools.json, /llms.txt.