Skip to content
Thryve

Local alternatives to popular AI products

Tools that run models on hardware you control instead of requiring a cloud chat provider.

LM Studio homepage showing local AI desktop app
LM Studio logo

LM Studio

A desktop application for discovering, downloading, running, chatting with, and serving AI models locally, with optional cloud features.

  • Hybrid
  • Partially self-hostable
  • Optional account
  • Works offline
  • Local AI
  • Desktop
Freemium
View privacy profile
Ollama homepage showing local large language models
Ollama logo

Ollama

A local model runner and API for downloading, managing, and running large language models on a user's device.

  • Hybrid
  • Open source
  • Self-hostable
  • Optional account
  • Works offline
  • Local AI
Freemium
View privacy profile
PrivateGPT GitHub project page
PrivateGPT logo

PrivateGPT

An open-source API layer for building private and self-hosted AI applications with local models, retrieval, tools, and document ingestion.

  • Self-hostable
  • Open source
  • No account
  • Works offline
  • Self-hosted AI
Open core
View privacy profile
Product photography and image-editing tools on a private desk
BUILT BY THRYVE

Thryve BG Remove

Remove image backgrounds with local AI, refine brushes, and replace backgrounds in the browser.

  • On-device
  • Optional account
  • Local AI
  • Web
Freemium
View privacy profile
Microphone, waveform transcript, and headphones for private transcription
BUILT BY THRYVE

Thryve Scribe

Transcribe audio and video with local Whisper or optional fast cloud mode.

  • Hybrid
  • Optional account
  • Local AI
  • Web
Subscription
View privacy profile
Microphone and notebook for private text-to-speech work
BUILT BY THRYVE

Thryve Speak

Turn English text into natural speech with local multi-voice AI (WAV).

  • On-device
  • Optional account
  • Local AI
  • Web
Freemium
View privacy profile

About this collection

Why this collection exists

Popular AI products often mean sending prompts, documents, or media to a remote model provider. This collection groups tools that can run inference or AI-assisted workflows on hardware you control, such as local runtimes, private model hosts, or first-party tools that keep optional AI processing under a clear on-device or user-controlled boundary.

Privacy context

Local AI reduces prompt and document exposure to a vendor model API, but it is not automatic privacy. Local tools still involve model provenance, telemetry, account options, and what leaves the machine during updates. Compare architectures and evidence rather than treating “local” as a pass/fail badge.

Inclusion criteria

Membership is curated against the rules below. It is not an automatic filter and does not grant review or verification status.

  • Primary AI or model workflow can run locally, on self-hosted infrastructure, or on-device without a mandatory third-party cloud model for core use.
  • Listings include local model runners, private GPT-style apps, and Thryve tools that expose local or clearly scoped AI paths.
  • Hybrid tools may appear when a documented local path exists; pure cloud-only AI assistants are excluded.
  • Model downloads, optional cloud models, and update checks may still use the network; those details stay on each privacy profile.