CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.text / smart-turn-v3
From: pipecat-ai <huggingface.co/pipecat-ai>Newsgroups: alt.freeweights.textSubject:

smart-turn-v3

Date: 2026-11-15Commit: f766f81Signed: 2DC28CC7B26A283B ✓

Voice activity detection model. Onnx weights, 91.0 MB in all. Published by pipecat-ai under BSD-2-Clause.

text91.0MBsmart-turn-v3onnxpermissive
FactsParamsnot statedSize91.0 MBFormatsonnxTaskvoice activity detectionLicenseBSD-2-ClauseFrom sourcepermissive
$ fw get pipecat-ai/smart-turn-v3
# manifest sha256 60e58e7bd5b6… · 91.0 MB · 12 files
$ fw verify ./smart-turn-v3 pipecat-ai/smart-turn-v3
From sourceHosted (signed)pipecat-ai/smart-turn-v3Last check: 2026-11-15 04:33 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Works with

Run2Convert3Scan1

Runs with

Ranked by fit to this model (onnx), then by GitHub stars.

UltralyticsOpenVINO

Convert with

ONNXOptimumOlive

Scan with

ModelScan
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
91.0 MB
Weights
8.68 MB (smallest of 5 ONNX variants)
Parameters
not stated
Formats
onnx
Task
voice activity detection
License
BSD-2-Clause (permissive)
Files
12, each with a sha256
Publisher
pipecat-ai
Added
2026-11-15
Section
alt.freeweights.text
Lineage

Original release from the publisher.

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 04:33 UTCpassed

Model card

From the publisher at commit f766f81. Shown as plain text with links; images open at the publisher.

Smart Turn v3.x#

Smart Turn is an open‑source semantic Voice Activity Detection (VAD) model that tells you whether a speaker has finished their turn by analysing the raw waveform, not the transcript.

Model architecture#

  • Backbone: Whisper Tiny encoder
  • Head: shallow linear classifier
  • Params: 8M
  • Checkpoint: 8 MB ONNX (int8 quantized), 32MB ONNX (unquantized)

How to use#

Please see the blog post and GitHub repo for more information on using the model, either standalone or with Pipecat.

Thanks#

Thank you to the following organisations for contributing audio datasets:

Files

12 files · 91.0 MBWeights 8.68 MB (smallest of 5 ONNX variants). Sizes and sha256 from the signed manifest.
Hide all 12 files with their sha256
FileSizesha256
.gitattributes1.52 KB11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361
README.md1.25 KB8b87ad1bb42432c1c8221944a7de5a0d1a52c2162af6f447ee9522530b38f199
benchmarks/smart-turn-v3.0.md5.20 KBa8d891975d4770d4b846e1b3cc25f6c30cc097f01dc58a277ed5b9535b856add
benchmarks/smart-turn-v3.1-cpu.md5.21 KBfd158ba597654a9fb7c9ca39ea68bcf1d08ba02b1edca1ea0bd8fd009c4ee088
benchmarks/smart-turn-v3.1-gpu.md5.21 KB5d24dca0a9875fa5fdabaa9f5ab5d2f3509b39a6e8fc6dfbace77f7d3ecca749
benchmarks/smart-turn-v3.2-cpu.md5.21 KB8d6c1175cb01f16df751569b304696b787b5dc148cedf445a97074743cef1928
benchmarks/smart-turn-v3.2-gpu.md5.21 KBd4ed84aead20fa148187c31d4f6a895f8cfece3d3a90e1f6dffe7fcc0e162ecb
smart-turn-v3.0.onnx8.76 MB07a133aba31e2d0b523f17f8c2e4e65efe6d8f685efd12ca4fe21ebf4e798991
smart-turn-v3.1-cpu.onnx8.68 MBfb68d55c2d542ce79e44b12013bfd571e90df8594ab096d757198e851b0c6594
smart-turn-v3.1-gpu.onnx32.4 MBa32f7445d5076029472b6c9f7a71005df576ea19d5f929021200f535b962af84
smart-turn-v3.2-cpu.onnx8.68 MB2bb026316b14a660486a75b1733cd3fbab8c2fd0314dc9af7be49f8cca967e4f
smart-turn-v3.2-gpu.onnx32.4 MBab8dc64b88713f90b571c15b714bd1330e6c883cad8763dacf65c9376dc539be
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
f766f81d3c pinned now2026-11-15 04:33 UTC12 files · 91.0 MB

Use it

$ fw get pipecat-ai/smart-turn-v3
# manifest sha256 60e58e7bd5b6… · 91.0 MB · 12 files
$ fw verify ./smart-turn-v3 pipecat-ai/smart-turn-v3
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/pipecat-ai/smart-turn-v3.json --dir ./smart-turn-v3 $ python3 fw.py verify ./smart-turn-v3 ./pipecat-ai__smart-turn-v3.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download pipecat-ai/smart-turn-v3 --revision f766f81d3cfdf7737ac64aad813d91bbfd56bf93 --local-dir ./smart-turn-v3
Run at home

License

Name
BSD 2-Clause License
Id
BSD-2-Clause
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
8b87ad1bb42432c1c8221944a7de5a0d1a52c2162af6f447ee9522530b38f199
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: pipecat-ai <huggingface.co/pipecat-ai>Newsgroups: alt.freeweights.textSubject: smart-turn-v3Date: 2026-11-15Commit: f766f81Signed: 2DC28CC7B26A283B ✓
Upstream
pipecat-ai/smart-turn-v3
Pinned commit
f766f81d3cfdf7737ac64aad813d91bbfd56bf93
Manifest
smart-turn-v3.json · .minisig
Manifest sha256
60e58e7bd5b6f6dea19aaa1082c2681023395111a119e64da0cec39987dbcb51
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 04:33 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/pipecat-ai/smart-turn-v3.json $ curl -O https://freeweights.allternit.com/manifests/pipecat-ai/smart-turn-v3.json.minisig $ minisign -Vm smart-turn-v3.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

1 model in the catalog name this one as their parent.

Quantizations
1
Quantizations 1
Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download pipecat-ai/smart-turn-v3
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. smart-turn-v3.json. Provenance. Takedown history: none.
Source
huggingface.co/pipecat-ai/smart-turn-v3

Variants

1 quantized or adapter build of smart-turn-v3. Each has its own license check and page.

SubjectSizeParamsLicenseDate
smart-turn-v3-ONNXHosted (signed) From: onnx-community in alt.freeweights.text94.7 MB—Permissive2026-11-15

Related models

All of alt.freeweights.text →
From: laion <huggingface.co/laion>Subject: a2-rl-e2egit_large-50-32B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-large-15-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-v2-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_pymethods2test-large-80-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_unitsyn-python-large-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_mix_h2_language_proportional-65-8B