Speech Model

OpenAI launches GPT-Live-1 API for full-duplex speech at $0.05 per minute: OpenAI has released GPT-Live-1 as a developer API, allowing applications to talk and listen simultaneously. The full-duplex speech model achieved an 80.1 percent score in interactivity tests, compared to 45.4 percent for its predecessor. The API is priced at $0.05 per minute.

AI Models • The Decoder

New inference-time method reduces hallucinations in OpenAI's Whisper: Researchers have developed a training-free, inference-time method to reduce hallucinated transcripts in OpenAI's Whisper model. The approach estimates a compact hallucination-associated subspace from non-speech calibration data and projects decoder hidden states away from it.

AI Research • arXiv

Microsoft debuts MAI-Transcribe-2 speech model with 72% price cut: Microsoft AI has debuted MAI-Transcribe-2, a speech recognition model that the company claims outperforms Gemini 3.5 Transcribe and GPT-Transcribe. The model is priced at an early-bird rate of $0.10 per audio hour through 2026, representing a roughly 72% price cut.

Microsoft AI • Techmeme