Speech recognition models
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 7 |
| Companies | 36 |
| Map changed | 2026-11-19 |
| Map | open the map |
Speech recognition models is a territory of the AI industry on map of ai. In short: speech recognition models convert spoken audio into text output. The map lists 36 products and companies in 7 categories.
papersthe last 30 days, "speech recognition models" in the title
- GigaSpeech 3: Scaling Multi-Dialect Arabic Speech Recognition with Weakly Labeled Data 2026-11-02
- Induction Head Origins of SLAM-ASR Speech Recognition Models 2026-11-18
- 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
- Sada-ASR: A Production Multidialectal Arabic Speech Recognition System 2026-11-14
- Earlybird: Predictive Streaming Speech Recognition in 29M Parameters 2026-11-10
- From Transcription to Reformulation: Defining and Benchmarking Grounded Generative Speech Recognition 2026-10-30
- Bio-Inspired Auditory Processing: Coupling Parallel Resonate-and-Fire Dynamics with Ternary Recurrent Networks for Speech Recognition 2026-10-21
The map of speech recognition models has 7 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
foundation asr models7 products and companies · its own article →
Foundation ASR models are base neural networks trained on large speech corpora to convert audio to text, used by developers and companies building speech applications. These models serve as starting points for fine-tuning on specific languages, accents, domains, or acoustic conditions.
Products and companies
Open ground
OpenAI, Meta, and Google release foundation ASR models with broad language coverage and open weights or APIs; none of them publishes real-time streaming-optimized foundation models with latency guarantees under 100ms on commodity hardware.
cloud speech apis8 products and companies · its own article →
Cloud speech APIs convert spoken audio into text and perform related voice processing tasks. Developers integrate these APIs into applications to add speech recognition, speaker identification, and voice command capabilities.
Products and companies
on-device and open source4 products and companies
On-device and open source speech recognition models that run locally without cloud dependency. Developers, privacy-focused users, and organizations integrate these into applications and devices.
Products and companies
transcription apps5 products and companies
Transcription apps convert spoken audio into written text using speech recognition models. Users include journalists, students, medical professionals, and business teams who need to document or search spoken content.
Products and companies
voice dictation4 products and companies
Voice dictation converts spoken words into text in real-time or near-real-time, used by professionals, students, and accessibility users to compose documents, emails, and messages without typing.
Products and companies: wispr flow, superwhisper, dragon, aqua voice.
meeting notetakers4 products and companies
Software that records meetings and converts spoken words to text in real time or post-recording. Used by business teams, legal firms, and remote workers to document discussions and create searchable records.
Products and companies: fireflies, fathom, granola, read ai.
medical scribes4 products and companies · its own article →
Medical scribes use speech recognition to transcribe clinical conversations and generate documentation during patient encounters. Physicians, nurse practitioners, and other clinicians use these systems to reduce administrative burden and create medical records.
Products and companies: abridge, nuance dax, suki, ambience healthcare.
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.