Teleprompter
FreeAn on-device AI for your meetings that listens to you and makes charismatic quote suggestions.
FreeFree tier
Inputs: audioOutputs: text
About Teleprompter
Teleprompter is an on-device AI tool that listens to your speech and suggests charismatic quotes to say in real time. It runs locally using semantic embeddings by default, with an optional highly experimental finetuned model. Currently limited to MacOS due to its reliance on the on-device hear speech recognition library. The tool requires Python, Transformers, and PyTorch and is distributed as an open-source GitHub repository under the MIT license.
Key Features
On-device speech recognition using the hear library (MacOS only)
Provides contextual quote suggestions to improve charisma
Default mode uses semantic embeddings for relevance
Optional finetuned model via command-line argument (experimental)
Open source under MIT license
Pros & Cons
Pros
- Works entirely on-device, preserving privacy
- Open source with an MIT license – free to use and modify
- Semantic embeddings provide relevant quote suggestions
- No dependency on cloud APIs or internet connectivity
Cons
- Only works on MacOS (uses the hear library for speech recognition)
- Requires manual environment setup (conda, transformers, pytorch)
- Finetuned model is highly experimental and may produce unreliable results
- Single-contributor project with limited maintenance (13 commits)
Best For
Enhancing charisma in live conversations and meetingsPracticing public speaking with real-time promptsGetting context-aware quote recommendations during discussionsExploring on-device AI speech interaction prototypes
FAQ
What is Teleprompter?
Teleprompter is an on-device AI tool that listens to your speech and provides charismatic quote suggestions in real time.
How do I run Teleprompter?
Run `python main.py` from the repository directory. By default it uses semantic embeddings; use `python main.py llm` to try the experimental finetuned model.
What dependencies does it require?
It requires Python, the Transformers library, PyTorch, and a general conda environment. These are listed in the repository's README.
Does Teleprompter work on Windows or Linux?
Currently it only works on MacOS because it uses the on-device `hear` library for speech recognition, which is MacOS-only.
Is the finetuned model available?
Yes, the finetuned model can be used by passing the `llm` argument, but it is marked as highly experimental.
Is Teleprompter free to use?
Yes, it is open source under the MIT license and completely free.