Technologist Mag
  • Home
  • Tech News
  • AI
  • Apps
  • Gadgets
  • Gaming
  • Guides
  • Laptops
  • Mobiles
  • Wearables
  • More
    • Web Stories
    • Trending
    • Press Release

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

What's On
‘Wicked’ Director Jon M. Chu on ‘What Makes Art Beautiful’ in the AI Era

‘Wicked’ Director Jon M. Chu on ‘What Makes Art Beautiful’ in the AI Era

17 December 2025
Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

17 December 2025
Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

17 December 2025
This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

17 December 2025
UPDATE: Dispatch Is Soaring To Switch 2 And Switch Next Month

UPDATE: Dispatch Is Soaring To Switch 2 And Switch Next Month

17 December 2025
Facebook X (Twitter) Instagram
Facebook X (Twitter) Instagram
Technologist Mag
SUBSCRIBE
  • Home
  • Tech News
  • AI
  • Apps
  • Gadgets
  • Gaming
  • Guides
  • Laptops
  • Mobiles
  • Wearables
  • More
    • Web Stories
    • Trending
    • Press Release
Technologist Mag
Home » Meta’s new open-source AI tool helps you clean up noisy recordings just by typing
Tech News

Meta’s new open-source AI tool helps you clean up noisy recordings just by typing

By technologistmag.com17 December 20253 Mins Read
Meta’s new open-source AI tool helps you clean up noisy recordings just by typing
Share
Facebook Twitter Reddit Telegram Pinterest Email
Meta’s new open-source AI tool helps you clean up noisy recordings just by typing

Cleaning up audio usually means scrubbing timelines and tweaking filters, but Meta thinks it should be as easy as describing the sound you want. The company has released a new open-source AI model called SAM Audio that can isolate almost any sound from a complex recording using simple text prompts.

Users can pull out specific noises like voices, instruments, or background sounds without digging through complicated editing software. The model is now available through Meta’s Segment Anything Playground that houses other prompt-based image and video editing tools.

🔉 Introducing SAM Audio, the first unified model that isolates any sound from complex audio mixtures using text, visual, or span prompts.

We’re sharing SAM Audio with the community, along with a perception encoder model, benchmarks and research papers, to empower others to… pic.twitter.com/FuMJyULmJR

— AI at Meta (@AIatMeta) December 16, 2025

Broadly speaking, SAM Audio is designed to understand what sound you want to work with and separate it cleanly from everything else. Meta says this opens the door to faster audio editing for use cases like music production, podcasting, film and television, accessibility tools, and research.

For example, a creator could isolate vocals from a band recording, remove traffic noise from a podcast, or delete a barking dog from an otherwise perfect recording, all by describing what they want the model to target.

How SAM Audio works

SAM Audio is a multimodal model that supports three different types of prompts. Users can describe a sound using text, click on a person or object in a video to visually identify the sound they want to isolate, or mark a time span where the sound first appears. These prompts can be used alone or combined, giving users fine-grained control over what gets separated.

meta-ai-sam-studio-sound-separation

Under the hood, the system relies on Meta’s Perception Encoder Audiovisual engine. It acts as the model’s ability to recognize and understand sounds before slicing them out of the mix.

To improve audio separation evaluation, Meta has also introduced SAM Audio-Bench, a benchmark for measuring how well models handle speech, music, and sound effects. It is accompanied by SAM Audio Judge, which evaluates how natural and accurate the separated audio sounds to human listeners, even without reference tracks to compare against.

Meta claims these evaluations show SAM Audio performs best when different prompt types are combined and can handle audio faster than real-time, even at scale.

meta-ai-sam-studio-multi-prompt

That said, the model has clear limitations. It does not support audio-based prompts, cannot perform full separation without any prompting, and struggles with similar overlapping sounds, such as isolating a single voice from a choir.

Meta says it plans to improve these areas and is already exploring real-world applications, including accessibility work with hearing-aid makers and organizations supporting people with disabilities.

The launch of SAM Audio ties into Meta’s broader AI push. The company is improving voice clarity on its AI glasses for noisy environments, working toward next-generation mixed reality glasses expected to arrive in 2027, and developing a conversational AI that could rival ChatGPT, signaling a wider focus on AI models that understands sound, context, and interaction.

Share. Facebook Twitter Pinterest LinkedIn Telegram Reddit Email
Previous ArticleBorder Patrol Bets on Small Drones to Expand US Surveillance Reach
Next Article The Very Best Gifts for Dads of Christmas 2025

Related Articles

‘Wicked’ Director Jon M. Chu on ‘What Makes Art Beautiful’ in the AI Era

‘Wicked’ Director Jon M. Chu on ‘What Makes Art Beautiful’ in the AI Era

17 December 2025
Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

17 December 2025
Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

17 December 2025
This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

17 December 2025
The Very Best Gifts for Dads of Christmas 2025

The Very Best Gifts for Dads of Christmas 2025

17 December 2025
Border Patrol Bets on Small Drones to Expand US Surveillance Reach

Border Patrol Bets on Small Drones to Expand US Surveillance Reach

17 December 2025
Stay In Touch
  • Facebook
  • Twitter
  • Pinterest
  • Instagram
  • YouTube
  • Vimeo

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Don't Miss
Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

Amazon launches Alexa+ on the web to take on ChatGPT and Gemini

By technologistmag.com17 December 2025

Amazon has begun rolling out a web version of its next-generation Alexa+ assistant, giving users…

Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

Google reveals Gemini 3 Flash to speed up AI search and beefs up image generation

17 December 2025
This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

This HP OmniBook X Flip sale gets you a Core Ultra 7 and 1TB SSD for less

17 December 2025
UPDATE: Dispatch Is Soaring To Switch 2 And Switch Next Month

UPDATE: Dispatch Is Soaring To Switch 2 And Switch Next Month

17 December 2025
The Very Best Gifts for Dads of Christmas 2025

The Very Best Gifts for Dads of Christmas 2025

17 December 2025
Technologist Mag
Facebook X (Twitter) Instagram Pinterest
  • Privacy
  • Terms
  • Advertise
  • Contact
© 2025 Technologist Mag. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.