Modulate, a frontier audio AI company, has secured $25m in fresh capital as organisations wake up to the fact that machines speaking fluently is only half the problem, and understanding what is actually said, and how, is the harder half.
Future Ventures led the round, joined by Hyperplane and Lakestar. The firm will direct the money into AI and machine learning research, product and engineering, developer relations and partnerships, while widening the range of APIs, models and deployment choices open to developers.
The raise arrives on the back of strong technical and commercial traction. Modulate’s models now process upwards of 10 million hours of audio monthly, and the company has passed the 600 million hour mark in total.
Its transcription and deepfake detection tools have each taken top position on Hugging Face benchmarks, with the firm claiming first place on the Open ASR Leaderboard and on the platform’s deepfake speech test. Transcription is priced at $0.03 per hour for batch jobs, and deepfake detection scores 98.9% accuracy on public benchmark data.
At the centre of the business sits Velma, a conversation understanding platform that the company says delivers twice the true positive accuracy of conventional LLMs and seven times fewer false positives.
Rather than working from text, Velma reads audio directly, picking up emotion, tone, intent, emphasis, synthetic speech and behavioural cues. These signals can stand alone or be combined to flag fraud attempts, AI agent failures, harassment, customer frustration and policy breaches, and Velma can act in real time so applications can step in mid-conversation.
Velma runs on Modulate’s Ensemble Listening Model (ELM) architecture, which coordinates more than 100 specialised audio models instead of leaning on one giant foundation model. According to the company, this approach has proven up to 1,000x more efficient than a single large model, cutting cost, energy and memory demands at scale.
Modulate builds audio-native AI that helps machines interpret voice conversations. Its technology shields healthcare institutions from deepfake attackers, improves emotional awareness in voice agents, curbs extremism and harassment online, tracks voice agent performance, identifies child grooming conversations and masks the voices of agents working in high-risk situations.
The company is also investing in its developer and partner push, building new industry models, adding SDKs and APIs, expanding developer relations, creating partner integrations and supporting fresh deployment environments. The goal is to let builders of voice agents, communications platforms and security tools embed advanced audio understanding without training specialised models in-house.
Modulate CEO and co-founder Carter Huffman said, “Voice is becoming a primary interface for AI, and that creates a whole new set of problems that can’t be solved from a transcript. We’re already using audio-native AI to protect organizations from deepfake attacks, help voice agents understand emotion and respond with more empathy, identify dangerous behavior in online conversations, and monitor whether voice agents are actually performing the way they’re supposed to.
“Underneath all of that are more than a hundred specialized models working together to understand what’s really happening across audio, with dramatically less cost and compute than traditional large models.”
Huffman added, “Developers shouldn’t have to rebuild the audio intelligence layer every time they create a new voice experience. Our mission is to build the models and infrastructure that let them focus on the application they want to create. The opportunity facing audio-native AI is expanding incredibly quickly. We’ve built the technology and proven it at scale, and this investment lets us grow the team and move faster to meet that demand.”
Copyright © 2026 FinTech Global
