Transcription

🗒Transcription 🎙️Podcast transcription ‍⚕️Medical transcription 📝Meeting transcription

MAI Transcribe 1.5

By Microsoft

MAI Transcribe-1.5 is Microsoft AI’s transcription model for accurate, fast, domain-specific speech-to-text across 43 languages.

NewMultimodal

Released 19d ago

🗒Transcription 🎙️Voice recognition 🎤Dictation

Ink 2

By Cartesia

Ink 2 is Cartesia’s speech-to-text model for voice agents, optimized for streaming transcription, semantic endpointing, noisy audio, and low-latency turn detection.

NewMultimodal

Released 30d ago

🗒Transcription 🔊Audio enhancement 🎙️Voice detection

Mega ASR

By Shanghai AI Laboratory

Mega-ASR is an open-source robust automatic speech recognition model built for difficult real-world audio with noise, far-field speech, obstruction, echo, distortion, and transmission dropout.

NewMultimodal

Released 1mo ago

🔍Image interpretation 🎥Video analysis 🤖Agents 🗒Transcription 📄Document analysis 🤔Logical reasoning

Nemotron 3 Nano Omni

By NVIDIA

Nemotron 3 Nano Omni is NVIDIA’s open multimodal reasoning model for agentic systems. It unifies text, image, video, and audio in a single efficient 30B-A3B hybrid MoE model, built to replace fragmented vision-language-audio stacks with one shared perception-and-context model for multimodal agents.

NewMultimodal

Released 1mo ago

🗒Transcription 🎙️Voice recognition 🎤Voice transcription

MiMo V2.5 ASR

By Xiaomi

MiMo-V2.5-ASR is Xiaomi MiMo’s state-of-the-art end-to-end ASR model for robust transcription across Mandarin, English, Chinese dialects, code-switching, songs, noisy audio, and multi-speaker conversations. It is an 8B model aimed at difficult real-world recognition scenarios rather than only clean speech.

NewMultimodal

Released 1mo ago

🎤Voice agents 📞Customer support 🗒Transcription

Grok Voice Think Fast 1.0

By xAI

Grok Voice Think Fast 1.0 is xAI’s fastest and most capable voice-agent model so far. It is built for real-time multilingual conversations, tool calling, and realtime search, and powers the same Grok Voice stack used in xAI’s apps and Tesla vehicles.

NewMultimodal

Released 1mo ago

🗒Transcription 🎙️Voice recognition 🎤Voice notes transcription 📝Meeting transcription

FunAudio ASR

By Alibaba

FunAudio-ASR, also called Fun-ASR, is Tongyi Lab’s end-to-end speech recognition model family for production ASR. It is trained on very large real-speech data, supports low-latency real-time transcription, and is designed for strong contextual understanding, domain terminology handling, and multilingual recognition.

NewMultimodal

Released 2mo ago

🗒Transcription 🎙️Voice recognition 📝Meeting transcription

Chorus v1

By Trelis Research

Chorus-v1 is Trelis’s open-weight two-speaker ASR model for short mono audio clips with overlapping speech. It is a fine-tune of Whisper Large v3 Turbo that outputs a separate timestamped transcript for each of two speakers, without requiring an upstream diarization step.

NewMultimodal

Released 2mo ago

🗒Transcription 🎙️Voice recognition 🎧Audio translation 🎵Music analysis

MOSS Audio

By OpenMOSS

MOSS-Audio is an open-source unified audio understanding model family from MOSI.AI, OpenMOSS, and the Shanghai Innovation Institute. It is built to handle speech, environmental sound, music, captioning, time-aware QA, and complex audio reasoning in one system, with 4B and 8B Instruct and Thinking variants.

NewMultimodal

Released 2mo ago

🗒Transcription 🎙️Voice recognition

MAI Transcribe 1

By Microsoft

MAI-Transcribe-1 is Microsoft’s multilingual speech-to-text model for high-accuracy production transcription. It is positioned as the most accurate transcription model in the world across 25 languages, with strong performance in noisy environments, resilience to accents and speaking styles, and speed fast enough for a wide range of real-world transcription use cases.

NewMultimodal

Released 2mo ago

🎤Dictation 🎙️Voice recognition 🗒Transcription

Atlas 1

By Willow

Atlas 1 is the voice AI model behind Willow, built for fast, high-accuracy dictation and polished speech-to-text writing workflows. It is designed to adapt to the user’s tone and speaking style in real time, turning natural speech into cleaned-up text for emails, messages, documents, and prompts.

NewAudio

Released 2mo ago

🗒Transcription 🌍Multilingual subtitle creation

Fish Audio STT

By Fish Audio

Fish Audio Speech-to-Text is Fish Audio’s transcription model and tool for podcasts, interviews, and conversational recordings. It converts audio to text with speaker labels, timestamps, and inline emotion or paralanguage tags, supports 100+ languages, accepts 24 audio and video formats, and exports transcripts as SRT, VTT, or JSON without requiring code.

NewMultimodal

Released 2mo ago

Gen 4 Mistral

Voxtral Mini 4B Realtime 2602

By Mistral AI

4B-parameter multilingual streaming ASR model for real-time transcription and translation across 13 languages, delivering sub-500 ms latency with open weights suitable for on-device or server use.

🗒Transcription 🌐Text translation

Audio

Released 4mo ago

Universal 3 Pro

By AssemblyAI

Universal-3 Pro is a promptable speech language model that turns raw audio into highly accurate transcripts, letting you control disfluencies, tags, speaker roles and code-switching through natural-language prompts.

🗒Transcription 🔊Text to speech 🌐Text translation 🔍SEO content

Audio

Released 4mo ago

Gen 4 Qwen

Qwen3 ForcedAligner 0.6B

By Alibaba

Multilingual forced alignment model that aligns speech and transcripts in 11 languages, predicting timestamps for arbitrary units in up to 5 minutes of audio with accuracy surpassing previous end-to-end aligners.

Audio

Released 4mo ago

Gen 4 FLUX

Flux 2

By Black Forest Labs

Flux 2 is the new version of the Flux UI kit for Livewire, adding native date pickers, calendars, charts, Tailwind 4 support, a strengthened free tier, and Laravel starter kits, all while staying tiny and dependency free.

👾Virtual companions 🌐Websites 🗒Transcription

Image

Released 6mo ago

🔊Text to speech 💬Chatting 🎥Videos 🗒Transcription

Dia 2

By Nari Labs

Dia2 is an open source streaming dialogue TTS model that generates speech in real time from partial text, supports audio conditioning for natural back and forth conversations, and ships 1B and 2B checkpoints under Apache 2.0.

Audio

Released 7mo ago

Gen 4 Eleven

Scribe v2 Realtime

By Eleven Labs

Scribe v2 Realtime is a low-latency speech model for live transcription and captioning. It streams partial and final text with stable timestamps, optional speaker labels, multilingual recognition, and clean JSON events for apps that need instant, accurate audio-to-text.

📜OCR 🖼️Image to text 🗒Transcription 🖼️Logos

Audio

Released 7mo ago

Gen 7 LFM

LFM2 VL 3B

By Liquid AI

LFM2-VL-3B is a 3B vision-language model that reads images with text and answers in natural language or structured JSON. It handles OCR, charts, tables, and screenshots with long context and low-latency streaming, making it practical for multimodal RAG and assistants.

Text

Released 7mo ago

Gen 3 LFM

LFM2-Audio-1.5

By Liquid AI

LFM2-Audio-1.5 is a unified audio foundation model for real-time speech, music, and sound effects. It transcribes, speaks, and generates or edits audio from natural prompts, supports cross-lingual use, returns stable timestamps, and streams low-latency output for production apps.

🔊Text to speech 🎵Music 🔊Sound effects 🗒Transcription

Multimodal

Released 8mo ago

Gen 4 Qwen

Qwen3 ASR Flash

By Alibaba

Qwen3-ASR-Flash is Alibaba Qwen’s all-in-one speech recognition model, built on Qwen3-Omni to stream low-latency transcripts in 11 languages, robust to noise, music and code switching, with optional text prompts to bias recognition.

Text

Released 9mo ago

🔊Text to speech 🎵Music 🗒Transcription

VibeVoice TTS 1.5B

By Microsoft

VibeVoice is Microsoft's open-source frontier voice AI family, unifying long-form ASR, multi-speaker TTS and low-latency streaming TTS using continuous speech tokenizers and next-token diffusion for efficient, high-fidelity speech over very long audio.

Audio

Released 9mo ago

Canary 180m flash

By NVIDIA

Canary 180M Flash is a compact automatic speech recognition model in NVIDIA’s Canary family, designed for fast speech to text transcription with a smaller parameter footprint.

Audio

Released 10mo ago

Gen 2

Canary 1B v2

By NVIDIA

Multilingual, multitask speech recognition model (ASR) in the NVIDIA NeMo family.

Audio

Released 10mo ago

Cabina transcriber v1

By Cabina

Cabina transcriber v1 is a real-time speech-to-text model with stable timestamps, punctuation, and optional speaker labels for meetings, calls, and streams.

Audio

Released 10mo ago

💬Chatting 🤖Task automation 💻Coding 🗒Transcription

Ming Omni

By InclusionAI

Ming-Omni is an end-to-end multimodal assistant that reads text, images, audio, and video, then replies with text or natural speech in real time. It supports tool calling, long context, and grounded answers.

Multimodal

Released 1y ago

Parakeet TDT

By NVIDIA

Parakeet TDT is NVIDIA’s low-latency, streaming speech-to-text model in the Parakeet family—a transformer-transducer architecture tuned for high accuracy, robust noise handling, and real-time partial transcripts, deployable on GPUs from edge to data center.

Audio

Released 1y ago

Gen 4 Eleven

ElevenLabs Scribe

By Eleven Labs

State-of-the-art transcription model supporting 99 languages; Features word-level timestamps, spe...

🔒Private conversations 🔊Text to speech 🌐Text translation 🗒Transcription

Audio

Released 1y ago

Gen 4 Doubao

Doubao Realtime Voice

By ByteDance

Doubao Realtime Voice is ByteDance’s low-latency voice interface for the Doubao models—streaming speech-to-text, LLM reasoning, and text-to-speech in a single session so agents can talk naturally, interrupt mid-sentence, and respond instantly on web or mobile.

Audio

Released 1y ago

Gen 4 MiniMax

MiniMax Speech 1.0 HD

By MiniMax

High-definition T2A-01 speech model with hundreds of preset voices, rich emotion control and rapid few-second voice cloning across many languages.

🔊Text to speech 🗣️Voice cloning 🗒Transcription

Audio

Released 1y ago

Moonshine Web

By Moonshine AI

In-browser speech recognition demo Space for Moonshine ASR, designed to run locally in the browser (WebGPU or WASM).

📜OCR 🖼️Image to text 🗒Transcription 🌐SEO

Multimodal

Released 1y ago

Gen 3 Mistral

Pixtral Large

By Mistral AI

Pixtral Large is Mistral’s flagship vision-language model. It takes images plus text and returns grounded, step-by-step answers—great for document OCR, charts/diagrams, UI screenshots, and general visual QA—with long-context support, tool/function calling, and reliable JSON outputs.

Text

Released 1y ago

Moonshine

By Moonshine AI

Moonshine's AI offers a voice interface that runs locally, so all your data remains 100% under your control.

Audio

Released 1y ago

Gen 4 CLOVA

CLOVA Voice EMR

By Naver

CLOVA Voice EMR is NAVER’s clinical speech solution for hospitals and clinics. It transcribes doctor–patient conversations and dictations in real time, inserts medical terminology correctly, and formats notes for EMR systems—reducing after-visit paperwork while preserving accuracy and privacy.

Audio

Released 1y ago

🎤Voice agents 🔊Text to speech 🗣️Voice cloning 🗒Transcription

Raon-Speech

By KRAFTON

Raon-Speech is KRAFTON’s open-source 9B bilingual speech language model for English and Korean. It is built as a unified speech AI system that supports speech-to-text, text-to-speech, spoken QA, speech chat, and text QA, with low-latency generation and optional speaker-conditioned TTS.

Multimodal

Released 2y ago

🗒Transcription 🎙️Voice recognition 🎙️Voice detection 🎤Voice commands

RealtimeSTT

By KoljaB

RealtimeSTT is KoljaB’s open-source low-latency speech-to-text library for real-time voice applications. It listens to microphone audio, detects speech activity automatically, supports optional wake-word activation, and performs fast transcription using Faster-Whisper, making it suitable for voice assistants and other live speech interfaces.

Audio

Released 2y ago