Notta Desktop · Bot-Free capture
Notta Desktop captures computer audio and your microphone without a Notta Bot joining; for local offline transcription, use the separate Pro Privacy Mode.
Conversations hold decisions, commitments, and context that are easy to lose. Notta turns permitted recordings into searchable knowledge, while you choose the capture and processing workflow that fits your requirements.
Searchable memory for teams that move fast
Capture → Understand → Visualize → Search → Act
Choose bot-free desktop capture, then choose cloud or local processing deliberately.
Get the consent required by law, contract, and workplace policy before you record. This matters in client work, education, research, and regulated settings. A bot-free workflow changes how audio is captured; it does not remove your duty to follow the rules for the meeting.
Quick definition
Notta is an AI-powered meeting productivity platform that turns online meetings, in-person conversations, and uploaded recordings into searchable, actionable knowledge. Notta Desktop is its local-first desktop transcription client: it adds computer-audio and microphone capture to desktop workflows, including Bot-Free recording on Mac and Windows. Standard Bot-Free recording uploads audio for transcription; Pro Privacy Mode is the separate option for local offline transcription.
That distinction matters for anyone handling confidential client calls, research interviews, internal reviews, or sensitive discussions. A transcript is only the start: when cloud processing is appropriate, Notta can help turn meeting content into summaries and other useful outputs.
See how bot-free meeting recording fits a desktop workflow, or review the separate Privacy Mode transcription option.
Choose the right path
“No bot in the meeting” describes the capture method. “Processed on this computer” describes where transcription happens. Choose each step with care.
Notta Desktop captures computer audio and your microphone without a Notta Bot joining. Audio is encrypted and uploaded for real-time transcription, so Bot-Free does not mean offline processing.
Best when a visible meeting participant is not wanted and cloud transcription is permitted.
Privacy Mode uses a local model to transcribe on the computer. Recordings and transcripts stay in the selected local workspace; cloud summaries, Notta Brain, sync, and normal team collaboration are paused or limited.
Best when local processing is needed and the device meets setup requirements.
Notta Memo captures in-person conversations and supported phone calls without a meeting bot, then stores audio on the device. Transcription and further AI work require syncing to Notta, so Memo is not local transcription.
For in-person capture, Memo has four MEMS microphones. Its bone-conduction microphone is for phone calls when Memo is attached to the back of a phone.
If you need local offline transcription, set up Privacy Mode rather than assuming an ordinary Bot-Free recording stays on the device.
A closer look
These product images illustrate the local workflow, from choosing Privacy Mode to reviewing a transcript. Images retain their full aspect ratio.
Practical outcomes
The goal is not to collect more recordings. It is to preserve the signal you need, with a clear path from capture to review and action.
Capture remote voices without a meeting participant. Desktop system audio and microphone input bring both sides of a supported call into the recording.
Choose the processing location on purpose. Use standard Bot-Free capture for cloud transcription or Pro Privacy Mode for local transcription.
Keep key details searchable. Review names, dates, amounts, specialist terms, and speaker labels against the recording before relying on them.
Preserve decisions and uncertainty. Separate confirmed decisions from proposals, and do not turn tentative language into a commitment.
Protect exported copies. Limit access to local folders and files, which may not have the same protections as the original workspace or sharing link.
Move from transcript to useful work. When cloud processing is permitted, meeting content can support summaries and deliverables that help a team act.
For client-facing work, start with client meeting capture that fits your engagement rules. Researchers can also review research interview transcription safeguards before recording.
Three steps
Set up the capture and processing path before a sensitive meeting starts. That small pause can prevent a wrong assumption about where the transcript goes.
Step 1
Check the applicable rules, then choose Bot-Free recording or Privacy Mode based on whether cloud processing is allowed.
You see: a clear capture and processing plan.
Step 2
Check microphone and system-audio access. For Privacy Mode, select a writable folder, download a supported offline model, and enable the mode before recording.
You see: permissions and mode ready to use.
Step 3
Capture or import permitted audio, transcribe locally if using Privacy Mode, then check important wording and export only what policy permits.
You see: a reviewed record ready for permitted use.
Privacy Mode needs internet for first sign-in, model setup, and periodic license validation; offline use is available between checks. Leaving Privacy Mode does not automatically upload its local files. A later cloud workflow requires deliberate export and reimport when permitted.
At a glance
Standard Bot-Free recording and Bot recording are separate options. Notta Web supports online meeting recording with a Notta Bot; Notta Desktop supports both Bot and Bot-Free recording. For secure meeting transcription, confirm which processing path your organization approves.
Know before you start
Local model coverage is not the same as Notta's cloud language support. Pick the model and operating system that match your use case before the meeting.
| Local transcription engine | Language coverage | Device or setup requirements |
|---|---|---|
| Default offline model | English, Japanese, Simplified Chinese, Cantonese | About 1 GB model storage, at least 4 GB RAM, and two CPU cores |
| Apple Speech Analyzer | English, Japanese, Simplified and Traditional Chinese, Cantonese, Korean, German, French, Spanish, Italian, Portuguese | Supported Mac app version and macOS 26 or later; select a specific language for file transcription |
| Computer-audio capture | Depends on the chosen transcription workflow | macOS 13 or later, or Windows 10 or later; permissions vary by system |
Mac permission note: Some Mac versions require screen-recording permission to access audio. Granting that permission does not itself record the meeting screen. Review the permission prompt and your organization's device policy before proceeding.
Evidence and safeguards
The useful question is not only whether a tool can record. It is where the audio and transcript go, what controls apply, and whether that workflow fits the work.
These controls do not establish that every account or workflow meets a particular engagement's requirements. Privacy Mode does not make every Notta workflow local or automatically meet every legal, healthcare, confidentiality, or institutional obligation. Privacy and compliance claims should be reviewed by the appropriate legal or security team.
“I just love the ability of Notta to record meetings, transcribe notes and then be able to go back to that and remember what I committed to. Very convenient and super helpful to review before my next meeting or when following up with tasks.”
Customer feedback about reviewing meeting records; it is not a claim about Privacy Mode.
For more on the product's approach to Notta security and privacy, review the requirements for your specific account and engagement before recording.
Compare workflows
Choose based on who joins the call, where transcription runs, and what you need afterward. “Depends” means the exact behavior varies by product or setup.
| Decision point | Notta Desktop | Bot-joined workflow | Basic local audio capture |
|---|---|---|---|
| Meeting participant | Bot-Free mode does not add a Notta Bot | A meeting Bot joins | No meeting Bot is part of the capture |
| Capture source | Computer audio and microphone | Notta Bot workflow; setup and joining are separate steps | Depends on the recorder and device |
| Local offline transcription | Available through separate Pro Privacy Mode | Not the local Privacy Mode workflow | Depends on the tool; local audio capture alone does not confirm transcription |
| What happens to the recording | Standard Bot-Free uploads audio; Privacy Mode stores it in the selected local workspace | Follow the selected service's processing and storage terms | Depends on the device and subsequent workflow |
The standard Notta Bot flow is distinct from Bot-Free capture: prepare the Bot setup before the meeting. When manually inviting a Bot, paste the meeting link into Notta after the meeting starts, then confirm that the Bot has joined and recording has begun. Bot-Free users start recording directly in Notta Desktop and do not paste a meeting link or invite a Bot.
Quick facts
These figures describe local setup options, not the broader cloud transcription service.
4
languages in the default offline model
~1 GB
storage for the default offline model
4 GB
minimum RAM for the default model
11
languages listed for Apple Speech Analyzer
Apple Speech Analyzer requires a supported Mac app version and macOS 26 or later. Default-model requirements are at least 4 GB RAM and two CPU cores. Check current eligibility, device support, and plan terms before setup.
Common questions
Bot-Free recording means Notta Desktop captures computer audio and your microphone without a Notta Bot joining the meeting. In standard Bot-Free mode, audio is encrypted and uploaded for transcription. Local offline transcription is a separate workflow: Pro Privacy Mode processes the recording on your computer.
No. Bot-Free describes how Notta Desktop captures meeting audio; standard Bot-Free recording uploads audio for transcription. Pro Privacy Mode is the separate local-processing option. In that mode, recordings and transcripts stay in the local workspace unless you deliberately export and later reimport material when permitted.
Computer-audio capture requires macOS 13 or later or Windows 10 or later. You also need to allow the required audio and microphone permissions for your operating system. Check both system audio and microphone inputs before an important call so you capture remote participants as well as your own voice.
The default offline model supports English, Japanese, Simplified Chinese, and Cantonese. On a supported Mac app version with macOS 26 or later, Apple Speech Analyzer supports English, Japanese, Simplified and Traditional Chinese, Cantonese, Korean, German, French, Spanish, Italian, and Portuguese. Local-model language support is separate from Notta's cloud language support.
Privacy Mode keeps its recordings and transcripts in the selected local workspace, but it does not make every Notta workflow local or satisfy every legal or institutional rule. Get required consent, follow your organization's policies, and confirm contract terms, retention, data location, and any healthcare or institutional requirements before use.
Privacy Mode eligibility includes Pro, Business, and Enterprise, subject to billing-method conditions. Invoice-billed accounts should confirm eligibility with Notta sales. Notta uses subscription plans rather than an official standard per-minute price, so check current plan details before choosing a plan.
For setup help, see the Notta Help Center. To learn about the product category, Notta Desktop is a desktop capture client; Bot-Free recording is only available in Notta Desktop, while Notta Web's online meeting recording uses a Notta Bot.
Capture with care. Keep the signal.
Start with Notta Desktop for Bot-Free capture, then use Pro Privacy Mode when local offline transcription is the right path for your work.
Check consent, plan eligibility, device permissions, and organizational requirements before recording.