News and site search

typhoon-whisper-large-v3-ctc

Word- and character-level **timestamps for Thai**: a 29M character-CTC head on the **frozen** `typhoon-ai/typhoon-whisper-large-v3` encoder (pinned revision `748e8a4`, recorded in `config.json`). One encoder pass gives transcription *and* alignment. It is **both an aligner and a compact ASR model**: it force-aligns given text (a transcript or ASR hypothesis) to word/character timestamps, and its greedy CTC decode transcribes directly (`--transcribe`, no text needed). `standalone_align.py` (bundled) has no other dependencies; the frozen encoder downloads automatically at the pinned revision. Files: `head.safetensors`, `config.json` (architecture + encoder id/revision), `ctc_vocab.json` (80-symbol Thai char vocab v1.0.0).

hayai-ocr-v2.5-nova

**Hayai OCR** is an ultra-lightweight (~150M parameter) vision-to-text model engineered for ultra-fast, crop-level transcription across **Japanese, Chinese, Korean, and English**. Hayai couples **Google's SigLIP2 NaFlex** vision encoder with a **custom 12-layer causal transformer decoder**. It transcribes dense, stylized, horizontal, and vertical text in a single forward pass without requiring an intermediate text-line detection stage (e.g., DBNet/YOLO). > **Note:** Hayai is designed specifically for **crop-level recognition**. For full-page scanning, pair it with a text detector.

parakeet-ctc-110m-coreml

NVIDIA's Parakeet-TDT-CTC-110M model converted to CoreML format for efficient inference on Apple Silicon. This is a hybrid ASR model with a shared Conformer encoder and two decoder heads: - **CTC Head**: Fast greedy decoding, ideal for keyword spotting - **TDT Head**: Token-Duration Transducer for high-quality transcription This will: 1. Download the original model from NVIDIA (`nvidia/parakeet-tdt_ctc-110m`) 2. Convert each component to CoreML format 3. Extract vocabulary and create metadata

wav2vec2-french-phonemizer

Fine-tuned facebook/wav2vec2-base-fr-voxpopuli-v2 for **French speech-to-phoneme** (without language model) using the train and validation splits of Common Voice v13. As this model is specifically trained for a speech-to-phoneme task, the output is sequence of IPA-encoded words, without punctuation. If you don't read the phonetic alphabet fluently, you can use this excellent IPA reader website to convert the transcript back to audio synthetic speech in order to check the quality of the phonetic transcription. The model has been finetuned on Commonvoice-v13 (FR) for 14 epochs on a 4x2080 Ti GPUs at Cnam/LMSSC using a ddp strategy and gradient-accumulation procedure (256 audios per update, corresponding roughly to 25 minutes of speech per update -> 2k updates per epoch)

distil-large-v3.5

Distil-Whisper is the knowledge-distilled version of OpenAI's Whisper-Large-v3, described in the paper Robust Knowledge Distillation via Large-Scale Pseudo Labelling. As the newest addition to the Distil-Whisper English family, Distil-Large-v3.5 maintains the high efficiency of its predecessors while delivering better performance. Compared to earlier models, Distil-Large-v3.5 has been trained on over 4× more diverse public data (98k hours) and uses a "patient" teacher with an extended training schedule and aggressive data augmentation (SpecAugment) during distillation. This results in enhanced robustness and accuracy compared to previous Distil-Whisper models, making it suitable as a drop-in replacement. 1. It offers a different balance between accuracy and efficiency, remains **~1.5x faster** than Whisper-Large-v3-Turbo while performing slightly better on short-form transcription and falling ~1% behind on long-form transcription. 2. It works perfectly as a draft model for **speculative decoding** with Whisper-Large-v3. By keeping the encoder frozen during training, we need to load just two extra decoder layers and forward the encoder only once. This achieves ~2x faster inference compared to Whisper-Large-v3 while maintaining identical outputs.

parakeet-tdt-0.6b-v3-coreml

On‑device multilingual ASR model converted to Core ML for Apple platforms. This model powers FluidAudio’s batch ASR and is the same model used in our backend. It supports 25 European languages and is optimized for low‑latency, private, offline transcription. The conversion script is here: https://github.com/FluidInference/mobius/tree/main/models/stt/parakeet-tdt-v3-0.6b/coreml And the benchmarks are here: https://github.com/FluidInference/FluidAudio/blob/main/Documentation/Benchmarks.md#transcription

LightOnOCR-3-1B-GGUF

> **LightOnOCR-3-1B**, developed by lightonai, is a highly performant, lightweight OCR and end-to-end document understanding model designed as a seamless, drop-in upgrade for existing LightOnOCR-2 deployments. Released under the Apache 2.0 license, it retains the proven LightOnOCR-2-1B architecture while introducing powerful new visual understanding capabilities, including a **grounding mode** (triggered by the `grounding` prompt) that outputs labeled bounding boxes for all document elements in normalized page coordinates, short descriptions for images, and structured HTML data extraction from charts. Alongside its default **transcription mode** that outputs full-page markdown text from an empty prompt, the model is highly versatile—effortlessly handling complex multi-column layouts, tables, receipts, forms, handwriting, and math notation—offering a ready-to-use, single-model alternative to traditional, multi-stage document processing pipelines when images are optimally preprocessed at 200 DPI.

voxtral-mini-4b-realtime-GGUF

GGUF quantizations of mistralai/Voxtral-Mini-4B-Realtime-2602, a **4.4B-parameter realtime streaming speech-to-text model** with a causal audio encoder and configurable transcription delay. ⚠ This table used to name the F16 `voxtral-mini-4b-realtime.gguf`, and no such file was ever uploaded. The published name is `…-f16.gguf`, matching the quant suffix CrispASR's `-m auto:f16` resolver expects.

parakeet-realtime-eou-120m-coreml

CoreML conversion of nvidia/parakeet-realtime-eou-120m-v1 for streaming speech recognition with end-of-utterance detection on Apple Silicon. Used by FluidAudio for real-time transcription. The RNNT pipeline is split into three CoreML models, exported at two chunk sizes:

transcription-fixer-gemma-3-12b-GGUF