Recording and recording is getting easier in meetings, interviews, classes and daily work, but the real difficulty is often not "recording" but ratherFind truly valuable information from a large number of audio and video。
AudioNotes is an open source local audio-video to text-to-text and smart note-taking tool. What it is trying to solve is not just the issue of "speech to text", but to establish a path fromAudio and video → Text → Structured Notes → AI Q & Acomplete workflow. The project source code is disclosed on GitHub, and its core feature is local priority and try to keep users 'data on their own computers.
1. What problems does AudioNotes solve?
The way traditional recording tools work is simple: record it, and then save the file.
The problem is that as more and more recordings are recorded,"preservation" itself becomes a new burden.
For example, at the end of a two-hour meeting, you might need to answer:
- Who proposed a plan?
- What was the final decision?
- Who is responsible for a certain task?
- What is the deadline?
- When was a certain point made?
- What are the most important conclusions of this meeting?
Without a transcript, looking for this information often means listening to the recording again.
AudioNotes 'idea is to first use the local speech recognition model to convert audio and video into text, and then use the local big language model to organize the content, finally forming Markdown notes, and allowing users to continue to ask questions about this content. Project README positions it as a tool suitable for scenarios such as meetings, interviews, courses, and voice memos.
Therefore, it is more like a "Audio and Video Knowledge Base Entrance", not just a recorder.
2. Complete workflow: from sound to knowledge
The usage process of AudioNotes can be summarized into four stages.
1. Get audio and video
Users can upload audio or video files directly, or they can use a browser to record.
This means that it is suitable for processing existing material and for direct use in scenarios such as meetings and interviews.
2. local speech recognition
AudioNotes uses by default FunAudioLLM/Fun-ASR-Nano-2512 Perform voice recognition.
Compared with simply uploading files to an online transcription service, the greatest value of local identification is that audio and video content can be left on the user's own device.
Of course, local identification does not mean that the results must be perfect. Recording quality, environmental noise, accents, and multiple people speaking at the same time may affect the final draft. Therefore, important content such as contracts and meeting resolutions should still be confirmed by listening to the original audio.
3. AI automatic collation
After getting the original script, AudioNotes uses the local large language model run by Olama to organize notes.
Project default use qwen3.5:2bThe goal is to achieve a balance between the quality of Chinese processing and resource consumption under the resource conditions of ordinary computers.
This step is very critical.
Because a complete meeting transcript may contain a lot of colloquial expressions, repetitions, and irrelevant information, what structured notes really need is:
Conclusions, opinions, tasks, time, people and context.
Therefore, the value of AudioNotes does not only come from ASR, but from the combination of ASR and LLM.
4. Continue to ask questions about the content
After the notes are generated, users can continue to have conversations with the audio and video content.
For example:
"What three plans were finally determined at this meeting?"
Or:
"What tasks are Zhang San responsible for?"
This has transformed audio and video from a document that requires manual "reading" to an information source that can be directly queried.
3. Local priority is the most noteworthy feature of AudioNotes
Today's AI products rely heavily on cloud APIs, but audio and video often contain very sensitive information.
An internal company meeting may contain a business plan; a user interview may contain personal information; and a course recording may involve research that has not yet been made public.
The design concept of AudioNotes is precisely where different choices are made.
The project README clearly states that audio and video, text scripts, notes and conversation history are saved locally by default; speech recognition is completed locally through FunASR, and the large language model is run locally through Olama. By default, audio and video content will not be sent to third-party AI interfaces. Web pages also allow only local access by default.
From an architectural perspective, this is a typical local-first AI Thought:
Data enters the computer → complete ASR locally → complete LLM processing locally → save the results locally.
This is particularly attractive in privacy-sensitive scenarios.
Of course,"running locally" does not mean absolute security. The computer's own accounts, disk permissions, backup and network environment still need to be protected. The project also specifically reminds users to use and properly protect local data on a trusted computer.
4. Docker makes deployment relatively simple
For ordinary users, the biggest threshold is often not using AI, but installing AI.
AudioNotes provides Docker methods that encapsulate the main operating environment.
The officially recommended basic environments include Docker Desktop and Olama, and it is recommended that Docker Desktop allocate at least 12GB of memory; if processing longer audio and video, 16GB is recommended, and about 10GB of disk space is reserved.
The basic process is:
ollama pull qwen3.5:2b
mkdir audionotes
cd audionotes
curl -fsSL https://github.com/harry0703/AudioNotes/raw/main/docker-compose.yml -o docker-compose.yml
curl -fsSL https://github.com/harry0703/AudioNotes/raw/main/.env.example -o .env
docker compose up -d --remove-orphans
Access local 15433 Port to enter the Web interface. When performing speech recognition for the first time, you also need to download about 2.15GB of speech models; after the download is completed, the model will be saved on the local machine, and there is no need to download repeatedly in the future.
It's worth noting that the project's Docker configuration does not simply put a Web service up and running. Its Compose configuration also includes settings such as data persistence, health check, automatic restart, process recycling, and container permission shrinkage. For example, by default, only services are bound to 127.0.0.1, and passed cap_drop: ALL and no-new-privileges Reduce the risk of containers being abused.
This shows that in addition to being "operational", the project also considers some security and stability issues in actual deployment.
5. Where is the data stored?
Another key issue with native AI tools is:Where did my data go?
When AudioNotes uses Docker, it maps the data to the created audionotes Folder, which includes uploaded files, model cache, logs, and conversation history. In the source code operation mode, personal data is mainly stored in the project's storage folder, and Ollama's models are managed by Ollama himself.
This brings a very practical advantage: data management is relatively intuitive.
For example, project recommendations:
- Copy entire when backing up
audionotesfolder; - Copy the folder as a whole when migrating a computer;
- When historical data is no longer needed, you can stop the service and delete the folder.
In other words, users do not have to rely on a third-party SaaS platform to manage their historical notes.
6. Who is it suitable for?
AudioNotes is best suited for thoseHave a lot of voice or video data, but want the content to remain localizedpeople.
meeting minutes
This is the most direct application scenario.
After the meeting, you don't need to spend a lot of time manually sorting out the recording. You can let the AI complete the preliminary transcription and structured organization first, and then manually confirm the key conclusions.
Interviews and Interviews
Interviews often involve a lot of natural language and long conversations. After converting the recording to text, it will be much easier to find people, opinions and key answers.
curriculum and learning
For long course recordings, transcripts and AI Q & A can reduce the cost of listening back.
For example, instead of relistening to an hour of class to find a certain concept, you might as well ask directly:
"What examples did the teacher give of this concept?"
Personal voice memo
Inspiration, to-do items, and temporary thoughts can all be saved through recording, and then let AI help organize them into easier to read text.
7. Advantages and limitations of AudioNotes
If evaluated from a product perspective, the advantages of AudioNotes are relatively clear.
First, local priority is given.
This is its core competitiveness that distinguishes it from a large number of cloud AI note-taking tools. Sensitive audio and video do not have to be uploaded to third-party AI services by default.
Second, the workflow is complete.
It does not just solve "speech to text", but further completes note sorting and content Q & A.
Third, open source and self-deployable.
Users can view the source code, deploy it themselves, and adjust the model and configuration as needed.
Fourth, data management is relatively transparent.
Upload files, historical data, and model caches all have clear local storage locations.
But it also has obvious thresholds.
First of all, local AI has requirements for hardware resources. The project clearly recommends allocating at least 12GB of memory for Docker, and 16GB for long audio videos.
Second, the speed and effectiveness of local models will be affected by computer hardware.
Third, although the local model has privacy advantages, it does not automatically guarantee transliteration accuracy. For important content, manual proofreading still cannot be omitted.
Finally, installing Docker, Ollama, models and processing model caching still has a certain learning cost for users with no technical background at all.
8. What's really interesting about it is not the "transliteration"
If you only regard AudioNotes as a "record-to-text software", you will actually underestimate it.
What really deserves attention is that it converts traditional audio and video files into a content carrier that can be understood and queried by AI.
Past:
Recording → File → Manual playback
Now:
Recording → ASR → Text → AI Summary → Conversational Query
This means that audio and video itself has begun to become a kind of "interactive data."
Especially as local models become stronger and consumer-grade computer computing power continues to increase, this model may become an important route in addition to traditional cloud AI tools.
9. Conclusion
AudioNotes is a representative native AI application: it does not attempt to reinvent a complex AI platform, but combines several mature technical components--FunASR is responsible for understanding the sound, Ollama is responsible for understanding the content, Markdown is responsible for carrying notes, and the Web UI is responsible for interaction。
The most commendable thing about it is that it puts "privacy" and "AI capabilities" in the same product design.
For people who regularly handle meetings, interviews, courses, or personal recordings, AudioNotes provides a way of working that is worth trying:
No longer treat recordings as files that need to be saved, but treat them as a source of personal knowledge that can be searched, summarized, and questioned.
Project source code and complete deployment instructions can be found in AudioNotes repository on GitHub Check. The functionality and deployment information in this article is based on the project's current README and Docker Compose configuration.