⌘K
Change language Switch ThemeSign In
Narrow Mode
Modulate AI Launches Velma: Real-Time Voice Deepfake Detection API
Modulate AI Launches Velma: Real-Time Voice Deepfake Detection API
 ### meng shao@shao__meng
Modulate AI @modulate_ai 发布语音深度伪造检测 API “Velma”，在 Hugging Face 排行榜上以 98.9% 准确率 位居第一 🏆modulate.ai/api/deepfake-d…5
传统检测方案通常仅在通话开头检查前 10 秒，通过后即默认全程安全。这种“单次门禁”模式被诈骗者轻易绕过：他们用真人声音（本人、同事或录音）开启通话，骗过初始检测，随后中途切换为 AI 克隆语音，导致后续诈骗成功。
Velma 的关键改进
· 全程实时监测：每 2 秒分析一次音频流，实时捕获中途语音切换，而非仅开头或随机抽查。
· 成本大幅降低：每小时仅 0.25 美元，较竞品（30-150 美元/小时）便宜约 120 倍，真正实现经济可行的全程监控。
· 低延迟：仅需 2.5 秒 音频即可检测，适合短片段和实时场景。
· 高性能：准确率领先，错误率低于体积大 10 倍的模型，支持实时流式与批量处理。Show More
#### Sumanth
@Sumanth_077 · 14h ago
Massive breakthrough in voice deepfake detection! @modulate_ai just released a deepfake detection API that topped @huggingface's leaderboard at 98.9% accuracy.
Here's the problem with how most companies handle deepfake detection. They check the first 10 seconds of a call. If it passes, they assume the whole call is clean. Gate check. One scan. Done.
Fraudsters know this. So they open the call with a real voice. Their own voice, a colleague, a quick recording. Pass the check. Then switch to the AI-generated clone mid-call. The system already gave them the green light. They're through.
The fix is obvious. Monitor the entire call. Not just the opening. Not random spot checks. Every segment, continuously, in real-time.
But that was too expensive. Until now.
Velma is Modulate's real-time and batch deepfake detection API.
Here's what changed.
• Real-time streaming detection. Analyzes audio every 2 seconds during live calls. Catches mid-call voice switches instantly.
• 120x cheaper than competitors. $0.25 per hour instead of $30-150. Now you can actually afford to monitor full conversations instead of spot-checking.
• Only needs 2.5 seconds of audio. Faster detection, works with short segments.
• 98.9% accuracy, ranked first on HuggingFace. Lower error rate than models 10x larger.
First 1000 API credits are free.
I've shared the link in the replies!Show More
7
7
18
2,556
Apr 6, 2026, 1:30 AM View on X
0 Replies
2 Retweets
3 Likes
827 Views  meng shao @shao__meng
One Sentence Summary
Modulate AI introduces the Velma API, addressing the 'gate check' vulnerability in traditional deepfake detection through real-time streaming monitoring and significant cost efficiency.
Summary
This tweet highlights 'Velma,' a real-time voice deepfake detection API launched by Modulate AI. Velma addresses the pain point of traditional solutions that only perform a 'gate check' at the beginning of a call. By performing real-time streaming analysis every 2 seconds, it can effectively capture mid-call voice switches. Additionally, the API offers excellent performance in terms of cost ($0.25/hour) and latency (2.5 seconds), achieving 98.9% accuracy on the Hugging Face leaderboard, making it highly practical.
AI Score
86
Influence Score 2
Published At Today
Language
Chinese
Tags
Modulate AI
Velma
Deepfake Detection
AI Safety
API HomeArticlesPodcastsVideosTweets