←Story detail

Trending story

Perplexity releases pplx-decider-flash-v1.2 as a Mac app

2 reports2 story updatesUpdated Oct 24 17:59

Understand this story

AI overview

Perplexity has packaged the pplx-decider-flash-v1.2 decision model as a Mac app and released it; users can run the model locally without an API key, and data does not leave the device. According to Perplexity Developers, this is a 4B decision model that can take questions about text or images and return probabilities; the app takes up 5.8 GB of space and requires 16 GB of memory. Aravind Srinivas reposted related content and said this is the power of open weights and compact models. Currently, there are no follow-up developments to this story, and its status remains at the release and local-running description.

AI-generated from public reports

Latest development · Report

Perplexity launches Mac local decision model app

Developments

1 confirmed developments

  1. Perplexity releases pplx-decider-flash-v1.2 as a Mac app

    Source:Testing Catalog

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Oct 24

  1. Testing Catalog

    Perplexity launches Mac local decision model app

    Perplexity has packaged the pplx-decider-flash-v1.2 decision model as a Mac app that can provide a decision API locally without an API key. The model takes up 5.8 GB of space and requires 16 GB of memory; according to Perplexity Developers, this is a 4B decision model that can answer questions about text or images and return probabilities.

Oct 23

  1. Aravind Srinivas

    Perplexity Decision Model can run locally on Mac

    Perplexity developers announced that users can now run pplx-decider-flash-v1.2 locally on a Mac, without an API key, and data will not leave the device. Aravind Srinivas reposted this and said it is the power of open weights and compact models.

Related stories

  1. Related storiesRecursive releases open-weight local coding model Flint
  2. Related storiesAnt Ling Open-Sources the Ming-3.0 Speech Series, Five Models in Total
  3. Related storiesDeepSeek releases DeepSeek-V4.1-Flash-1017
  4. Related storiesMeta releases open-weight model Muse Spark Open