Speech audio foundation models
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | territory of the AI industry |
|---|---|
| Categories | 14 |
| Companies | 80 |
| Part of | neural networks |
| Map last changed | 2026-08-29 |
| Map | open the map |
Speech audio foundation models is a territory of the AI industry on map of ai. In short: speech audio foundation models are neural networks trained on large audio corpora to learn acoustic and linguistic abstractions, used for speech recognition, translation, speaker identification, and audio generation tasks. The map lists 80 products and companies in 14 categories.
The map of speech audio foundation models has 14 categories. Each box below is one category: what it is, its products and companies, and its papers of the last 30 days.
automatic speech recognition12 products and companies · its own article →
Automatic speech recognition converts spoken audio into text transcripts. Developers, enterprises, and end-users integrate ASR into applications for voice commands, transcription, accessibility, and call center automation.
Open ground
OpenAI Whisper, Google, Amazon, and Microsoft each ship multilingual ASR that handles accented speech and background noise; none of them ships a commercial ASR system that reliably transcribes speech in code-switched or heavily accented English at the same accuracy as monolingual standard English.
neural text-to-speech9 products and companies
Neural text-to-speech systems convert written text into spoken audio using deep learning models trained on large speech corpora. Publishers, accessibility teams, customer service platforms, and content creators use these to generate voice-overs, narration, and conversational audio at scale.
speech-to-speech / voice llm7 products and companies
Products and companies
self-supervised speech encoders5 products and companies
Self-supervised speech encoders are neural networks trained on unlabeled audio to learn general-purpose speech representations, used by researchers and engineers building downstream speech applications like ASR, speaker identification, and emotion detection.
Products and companies
Papers, last 30 days
audio-language models5 products and companies
Audio-language models process and generate spoken language by learning from speech audio paired with text or other modalities. Researchers, speech recognition teams, and voice interface developers use them to build systems that understand and produce natural speech.
Products and companies
voice cloning & persona5 products and companies
Products and companies
speech translation4 products and companies · its own article →
Speech translation converts spoken audio in one language directly to speech or text in another language, used by conference attendees, multilingual teams, and content creators to understand foreign-language speakers in real time.
Products and companies
Papers, last 30 days
- Machine interpreting competitiveness: a discourse-sensitive comparison of meaning and target-language quality in English-Chinese simultaneous interpreting 2026-11-15
- Walrus Technical Report 2026-11-10
- Text Grouping Changes Completion-Time Comparisons in Long-Form Simultaneous Speech Translation 2026-11-09
- Holistic Parallel Supervision for Expressive Speech to Speech Translation 2026-11-04
- CRESiST: Self-Enhancing Explorative Policy Learning for Simultaneous Speech Translation 2026-11-04
speech emotion & paralinguistics4 products and companies
Products and companies
audio generation & sound synthesis4 products and companies
Products and companies
speech foundation models5 products and companies
Speech audio foundation models are large neural networks trained on unlabeled speech data to learn general-purpose acoustic representations. Researchers and product teams use them as starting points for downstream tasks like recognition, translation, and speaker identification.
Products and companies
voice agents platforms6 products and companies · its own article →
Voice agents platforms enable developers to build, deploy, and manage conversational AI systems that interact with users through spoken dialogue. Enterprises, contact centers, and application developers use these platforms to create customer service bots, internal assistants, and voice-enabled applications.
Papers, last 30 days
- When Should a Voice Agent Speak? End-of-Turn Detection in Turkish 2026-11-19
- Does the Voice Matter? Evaluating Frustration Detection on Deployed Clinical Voice-Agent Calls 2026-11-16
- Latency, Speech Recognition and Turn-Taking in Voice Agents for Spanish Phone Negotiation: An Externally Measured Comparison of Pipecat, LiveKit and GPT-Live 2026-11-14
- Joi-Bench-Task: Benchmarking Task Coordination in Full-duplex Voice Agents through Interactive Simulation 2026-11-09
transcription and meeting apps5 products and companies · its own article →
Software and services that record, transcribe, and summarize spoken meetings in real time or post-recording. Teams, enterprises, and individuals use these to create searchable records and extract action items from calls and conferences.
Products and companies
music and sound generation5 products and companies
Music and sound generation systems that create audio from text, MIDI, or other control signals, used by musicians, producers, sound designers, and content creators to compose, arrange, and produce music and sound effects.
Products and companies
medical speech scribes4 products and companies
Medical speech scribes convert clinician-dictated or patient-spoken audio into structured clinical documentation, often with disease-specific templating. Hospitals, outpatient practices, and EHR vendors use them to reduce manual documentation burden during or after patient encounters.
Products and companies
see alsoadjacent territories with an article
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.