Back to News
IndustryAI Understanding briefing

Modulate raises $25 million to expand audio‑AI moderation and detection tools

Somerville‑based Modulate announced a $25 million funding round led by Future Ventures, aiming to grow its Velma audio moderation platform and new AI music detection model.

4 min readRead the linked source
Source-provided image accompanying Modulate raises $25 million to expand audio‑AI moderation and detection tools
Source referenceSource recorded
Publisher
digitalmusicnews.com
Source link
digitalmusicnews.comhttps://www.digitalmusicnews.com/2026/09/30/modulate-funding-round/
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Start here

Key terms

Large Language Model (LLM)
A language model trained on massive text corpora to generate and analyze text.
Ensemble
Combining predictions from multiple models to improve robustness or accuracy.
Compute
The processing resources required to train and run models, often measured in FLOPS or GPU hours.
Test yourselfAI Models Explained Quiz

What happened

Modulate secured $25 million in a new financing round led by Future Ventures, with participation from existing investors Lakestar and Hyperplane.

According to a report by Digital Music News, Modulate announced a $25 million financing round on September 30, 2026. The round was led by Future Ventures, a firm founded by Steve Jurvetson, and included participation from Lakestar—who led Modulate’s $30 million Series A in August 2022—and Hyperplane. The company, founded in 2017 and headquartered in Somerville, Massachusetts, said the new capital will be used to increase investment in AI/ML research, build new industry models, grow its engineering team, and expand its existing product portfolio.

Modulate’s flagship product, Velma, is described as an “ Listening Model (ELM) architecture that’s 2×‑4× more accurate than large language models (LLMs).” Velma is marketed as a real‑time moderation system that monitors customer‑support lines, video‑game voice chats, and other audio streams for toxic language, deepfakes, disruptive behavior, and emotional cues. The company cites a case where the game *Rainbow Six Siege* reduced critical toxic voice chat by 50% after deploying Velma.

In addition to moderation, Modulate offers a transcription service and a “Velma Triage” product. In June 2026 the firm launched an AI music detection model that provides segment‑level probabilities indicating whether vocals or instrumentals were AI‑generated. The tool is designed to generalize across different music generators and to identify broader AI‑generation patterns.

Source details: digitalmusicnews.com ↗

Why it matters

The capital will fund further AI/ML research, team expansion, and product development for Modulate’s audio‑focused models, which are already being used to curb toxic voice chat in games and detect AI‑generated music.

Audio‑focused AI tools are a growing niche with direct implications for online safety, content moderation, and intellectual‑property protection. Modulate’s claim that Velma is significantly more accurate than LLM‑based approaches suggests a potential shift toward specialized audio models that can operate with lower costs, which could lower barriers for smaller platforms to adopt advanced moderation.

The music‑detection model addresses a rising concern over AI‑generated audio infringing on copyrighted works. By offering segment‑based detection that is not tied to any single generator, Modulate positions itself to serve record labels, streaming services, and rights‑management firms that need scalable tools to monitor large audio libraries.

The funding round underscores investor confidence in audio‑AI as a commercial market. With backers like Future Ventures and Lakestar, Modulate joins a wave of AI startups receiving multi‑digit investments for niche modalities, indicating that capital is flowing beyond text‑centric models.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Model Parameter Size:8B Parameters
VRAM Required5.5 GBGPU memory footprint
Target HardwareMacBook / Single GPUDeployment tier
Privacy100% Air-GappedLocal device capability
Core takeaway: Small, quantized models (3B–8B) now run directly inside smartphones and laptops with complete data privacy, while mammoth 400B+ models remain the domain of datacenter clusters.
Interactive Concept Check+10 Points
AI Models Explained Quiz

In AI, what are a model's "parameters"?

What to watch next

Modulate’s rollout of its Velma and music‑detection services, adoption by gaming and music platforms, and competitive moves by other audio‑AI firms.

Adoption of Velma by additional gaming titles, live‑streaming platforms, and call‑center providers will reveal how quickly the technology scales beyond the initial case study.

The rollout of the music‑detection model to major record labels or streaming services could set industry standards for AI‑generated content monitoring and may trigger policy discussions around copyright enforcement.

Competitive responses from other audio‑AI firms—such as Stability AI’s recent $76 million Series B for audio projects—will indicate whether the market is consolidating around a few large players or diversifying across specialized solutions.

Potential regulatory scrutiny around deep‑fake audio detection and privacy implications of real‑time emotion analysis could affect product features and deployment strategies.

Related guides & quizzes

AI Models ExplainedAI EthicsFuture of AITest what you know — try a free AI quizLook up an AI term in our glossaryFollow the AI funding tracker
Found this useful?