Open source & local AI / Best open-source and local AI voice and text-to-speech tools
Best open-source and local AI voice and text-to-speech tools
3 tools for text to speech and voice cloning on your own computer. Your data stays on your machine, and most work offline once the model is downloaded. Licenses last verified 2026-10-24.
| Tool | Open source / local | Price | Notes | Checked | |
|---|---|---|---|---|---|
| Kokoro TTS Small open text-to-speech model (82M parameters) for local voiceovers | Open weights: Apache 2.0 | Free open weights (Apache 2.0) ✓ checked | 82M-parameter text-to-speech model | source 2026-10-24 | Get Kokoro TTS → |
| Piper TTS Fast local text-to-speech engine that runs offline | Open source: GPL-3.0 | Free (open source, GPL-3.0) ✓ checked | The earlier MIT-licensed repo was archived in 2025; this is the maintained one, and the project is looking for maintainers. Check the license of each voice you use | source 2026-10-24 | Get Piper TTS → |
| Chatterbox Open-source text-to-speech with voice cloning from a reference clip, 23+ languages | Open source: MIT | Free (open source, MIT) ✓ checked | Text-to-speech with voice cloning from a reference clip | source 2026-10-24 | Get Chatterbox → |
Where to run it
Open-source models need compute. Your own computer is the cheapest way to start. If it is too small, these are two providers to look at. We have not tested either one.
Disclosure: some links on this page are, or may become, affiliate links. If you buy through one we may earn a commission at no extra cost to you. It never changes which tools we list or how they rank. Details.
- Your own computer: small models run on a recent laptop; image and video models want a GPU or lots of memory.
- GPU cloud, pay as you go: Runpod rents GPUs by the hour for larger models and image generation. This is a referral link: if you sign up and use it, Runpod gives us platform credits (not cash) and may give you a credit bonus after you add funds.
- A server for self-hosted apps: DigitalOcean rents cloud servers for tools such as n8n, Open WebUI or Dify. Plain link, no affiliate relationship yet.
Check each provider's current pricing and referral terms before you commit.
Want it hosted or easier? Paid options
Self-hosting and free plans take setup time or have limits. If you would rather pay for a managed service, these fit the same jobs. Sponsored links are marked, and we show the free tier next to each price.
Disclosure: some links on this page are, or may become, affiliate links. If you buy through one we may earn a commission at no extra cost to you. It never changes which tools we list or how they rank. Details.
| Tool | Paid entry plan | Free tier | |
|---|---|---|---|
| ElevenLabs Realistic voiceover and voice cloning | Starter $5/mo ✓ verified | Yes: 10k credits/month; no commercial license | See ElevenLabs plans → |
| Murf Studio voiceovers for e-learning and ads | Creator $19/mo billed yearly ✓ checked | Trial only: 10 min voice generation; preview only, no downloads | See Murf plans → |
| Speechify Listening to text read aloud | Premium $29/mo ✓ verified | Yes: 10 robotic-sounding voices, up to 1.5x speed, text to speech only | See Speechify plans → |
Tell us your project and we will pick open-source and local tools for it.
Find local tools for my project“Open source” means the code is published under the license shown; “open weights” means the model files can be downloaded and run on your own hardware (the maker's chat app may still be a cloud service). LM Studio is free and runs models locally but is not open source. n8n and Dify are source-available: you can read and self-host the code, but their licenses are not OSI-approved open source. Running models locally needs a reasonably modern computer; larger models need a GPU or lots of memory. Licenses were read on the source linked on the check date; check the model's own license before commercial use. Some links are affiliate links (disclosure); they never change what we list.