
Impulse
A local AI workspace for Windows and Linux that runs models on your machine and answers from your own documents.
✓ Verified against veloresearch.com · Nov 11, 2026Free tierWindows · Linux

What it does
Impulse is an AI tool for running AI models locally, asking questions about your documents and chatting with an AI assistant, made by Velocity. It has a free plan.
Impulse is a local AI workspace from Velocity for Windows and Linux. It runs language models on the user's own machine, keeps documents on disk, and brings chat, model management, studio tools, apps, and system information together. Local inference means the documents used for questions do not need to leave the computer. The product's Momentum memory prepares a library of material on disk and reads it with the model's own tokenizer. It does not use an embedding model or a vector database. When a user asks a question, Momentum searches the library, sends only the relevant passages to the model, and cites those passages in the answer. Velocity reports benchmarking across 32 million tokens, configurable up to 1 billion. An OpenAI-compatible local endpoint lets other applications call the model on the same machine. Impulse is free for personal use, and commercial licensing is available for business use. It suits individuals and teams that want local model inference and answers based on their own documents, as well as developers who need a local OpenAI-compatible endpoint. Commercial terms are scoped to the use case and agreed with Velocity.
- Local inference on Windows and Linux
- Momentum document memory
- Passage citations in answers
- No embedding model or vector database
- OpenAI-compatible local endpoint
- Chat, Models, Studio, Apps and System workspace
- Free personal-use plan
- Commercial licensing
Verified facts
Each row links to the page we checked it on.| API | Yes, OpenAI-compatible local endpoint | veloresearch.com | ✓ Nov 11, 2026 |
|---|---|---|---|
| Commercial use | Yes, commercial licensing available | veloresearch.com | ✓ Nov 11, 2026 |
| Free plan | Yes, free for personal use | veloresearch.com | ✓ Nov 11, 2026 |
| Maker | Velocity | veloresearch.com | ✓ Nov 11, 2026 |
| Platforms | Windows and Linux | veloresearch.com | ✓ Nov 11, 2026 |
| Pricing | Free for personal use plus commercial licensing | veloresearch.com | ✓ Nov 11, 2026 |
What you can do with it
Good at
- Runs inference and keeps documents on the local machine
- Answers include citations to the passages used
- Momentum memory covers document sets without embeddings or a vector database
- OpenAI-compatible local endpoint for other software
Watch out for
- Windows and Linux only; no macOS build is listed
- Free use is limited to personal use; commercial use requires a licensing agreement
- Passages sent to an external model are subject to that provider's data handling
Common questions
- What is Impulse used for?
- Impulse is used for running AI models locally, asking questions about your documents and chatting with an AI assistant. On AIToolMaps it ranks #1 of 9 for running AI models locally.
- Is Impulse free?
- Yes. Impulse has a free plan.
- What platforms does Impulse run on?
- Impulse runs on Windows and Linux.
- Does Impulse have an API?
- Yes. Impulse offers an API, so developers can use it from their own code.
- Who makes Impulse?
- Impulse is made by Velocity.
- What are the best alternatives to Impulse?
- The closest alternatives are OfflineVault, LM Studio and Elicit, which do most of the same jobs. OfflineVault is the closest one with a free plan.
Reviews
No reviews yetListing history
- Nov 11, 2026Checked against veloresearch.com
- Nov 11, 2026Listed