{"slug": "hume-launches-the-first-decision-model-for-voice-ai", "title": "Hume launches the first decision model for voice AI", "summary": "Hume launched an Audio Decision Model that returns a probability for each possible answer to a developer-submitted question about a recording, and the company reports it agrees with a panel of human raters 70.7% of the time versus 56.5% for a single human rater. Hume said the model averages 0.82 across 10 benchmarks including MMAU, ParaPair and EmphAssess at 0.23 seconds median latency per screen, scoring higher and answering faster than Gemini 3.8 Flash, Qwen3.8-Omni-Flash, Inkling-Small, gpt-audio-1.5, Gemini 3.5 Flash Lite and Nemotron 3 Nano Omni. Hume cited 1B client hours processed and more than 10M human judgments across its work, and positions the Decision Model API as a decision layer in voice-agent systems alongside speech recognition and generation.", "body_md": "# Decision models for voice AI\n\nThe bar is human. Hume diagnoses system breaks and feeds failures into improvement.\n\nA sample telehealth call plays while three questions about it are answered for every five seconds of audio.\n\n## Why audio needs a decision model\n\nVoice AI considers tone, pauses, and sounds. Developers set Q&A; the audio decision model gives probabilities for quick audio judgments.\n\n## Different questions. The same audio.\n\nAsk objective, subjective or perceptual questions about speech and sound. Select a tab to explore examples.\n\nOne speaker\n\nTwo speakers\n\nThree or more speakers\n\n## Find the audio worth acting on\n\n### Ask more of the same audio\n\nApply several questions to the same recordings, from naturalness and clarity to empathy and timing.\n\n### Designed for fast evaluation\n\nRepeat evaluations across your collection as your model or criteria change.\n\n### Pinpoint cases for your next step\n\nFind candidate examples for training, isolate evaluation failures, and surface cases for escalation review.\n\n## More reliable than \n\na single listener\n\nThe Audio Decision Model agrees with the human crowd 70.7% of the time, while a single human rater agrees 56.5% of the time.\n\n- Number of client hours processed.\n- 1B hours\n- Human judgments across Hume’s work.\n- 10M+\n\nAgreement with a panel of human raters: Audio Decision Model 70.7%, Individual Rater 56.5%.\n\nAccuracy against speed: Hume’s Decision Model averages 0.82 over 10 leading benchmarks at 0.23 seconds median latency per screen, scoring higher and answering faster than Gemini 3.8 Flash, Qwen3.8-Omni-Flash, Inkling-Small, gpt-audio-1.5, Gemini 3.5 Flash Lite and Nemotron 3 Nano Omni.\n\nThe Audio Decision Model answers audio questions faster and more accurately than any other model on the market.\n\n- Average score over 10 leading benchmarks: MMAU, ParaPair, EmphAssess, etc.\n- 0.82\n- Median latency per screen\n- 0.23 s\n\n## Compare the judgments that matter\n\nAccuracy across the criteria you care about.\n\n1. **Hume’s Decision Model** Hume82%\n2. **Gemini 3.8 Flash** Google78%\n3. **Qwen3.8-Omni-Flash** Alibaba76%\n4. **Inkling-Small** Thinking Machines67%\n\n## Questions \n\nabout the model\n\nIt’s Hume’s audio-native decision model. Developers submit audio, a question and possible answers; it returns a probability for each answer. It helps teams evaluate speech and sound, assess how a voice comes across to listeners, and inform decisions in voice applications.\n\n## Audio decisions that fit your stack.\n\nSend audio, your question and possible answers. \n\nThe audio decision model returns a probability for each\n\nUse the Decision Model API as a decision layer in a voice-agent system with speech recognition and generation.", "url": "https://wpnews.pro/news/hume-launches-the-first-decision-model-for-voice-ai", "canonical_source": "https://www.hume.ai/products/decision-model-api", "published_at": "2026-11-02 17:07:10+00:00", "updated_at": "2026-11-02 17:22:49.349391+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-tools", "natural-language-processing"], "entities": ["Hume", "Audio Decision Model", "Gemini 3.8 Flash", "Qwen3.8-Omni-Flash", "Inkling-Small", "gpt-audio-1.5", "Nemotron 3 Nano Omni", "Google"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/hume-launches-the-first-decision-model-for-voice-ai", "markdown": "https://wpnews.pro/news/hume-launches-the-first-decision-model-for-voice-ai.md", "text": "https://wpnews.pro/news/hume-launches-the-first-decision-model-for-voice-ai.txt", "jsonld": "https://wpnews.pro/news/hume-launches-the-first-decision-model-for-voice-ai.jsonld"}}