# parakeet-tdt-0.6b-v2 by MLX Community: Open-Weight Model Source: https://savrn.com/models/parakeet-tdt-0-6b-v2 Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve parakeet-tdt-0.6b-v2 (618M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 1.2 GB | 1.5 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.6 GB | 0.7 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.3 GB | 0.4 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 24, 2026. [parakeet-tdt-0.6b-v2 on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/parakeet-tdt-0-6b-v2/gpus) ## SAVRN's Notes on parakeet-tdt-0.6b-v2 This is not the original Parakeet — it is the MLX conversion, rebuilt from nvidia/parakeet-tdt-0.6b-v2 for Apple silicon. At 618M parameters the working set is 1.5 GB at 16-bit and 0.4 GB at 4-bit, and the Index's cheapest listed setup of one MI300X at $1.85 an hour is a datacenter answer to a build that is meant to run in an M-series chip's unified memory. CC BY 4.0 permits commercial use with credit and an indication of changes, a lighter grant than some speech models carry. The tradeoff is ecosystem: MLX locks you to Apple hardware and the mlx-lm tooling, so if you are on CUDA you are better served by the upstream NVIDIA repository this was converted from. It draws 1.1M downloads a month; for batch transcription on a Mac, this is the build to reach for. SAVRN Research ## Model Card By MLX Community, published under cc-by-4.0, revision 8ae155301e23. This model was converted to MLX format from nvidia/parakeet-tdt-0.6b-v2 using the conversion script. Please refer to original model card for more details on the model. Read MLX Community's full model card This model was converted to MLX format from [nvidia/parakeet-tdt-0.6b-v2](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) using [the conversion script](https://gist.github.com/senstella/77178bb5d6ec67bf8c54705a5f490bed). Please refer to [original model card](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) for more details on the model. ### Use with mlx #### parakeet-mlx ``` pip install -U parakeet-mlx ``` ``` parakeet-mlx audio.wav --model mlx-community/parakeet-tdt-0.6b-v2 ``` #### mlx-audio ``` pip install -U mlx-audio ``` ``` python -m mlx_audio.stt.generate --model mlx-community/parakeet-tdt-0.6b-v2 --audio audio.wav --output somewhere ``` ## Identity and Version Repository mlx-community/parakeet-tdt-0.6b-v2 Publisher MLX Community Task Speech recognition Modality Audio Library mlx Parameters 618M parameters Languages mlx Revision 8ae155301e23d820d82aa60d24817c900e69e487 First published 2025-05-06 Last updated 2025-05-10 ## Files and Weights 7 files, 2.5 GB in total. The weights are 1 file totalling 2.5 GB in safetensors. Weights1 file · 2.5 GB Configuration1 file · 36.2 KB Tokenizer3 files · 266.9 KB Documentation1 file · 945 B Repository1 file · 1.5 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 2.5 GB | b958c37a6baa | | config.json | Configuration | 36.2 KB | — | | README.md | Documentation | 945 B | — | | .gitattributes | Repository | 1.5 KB | — | | tokenizer.model | Tokenizer | 251.4 KB | 5a3a82c48998 | | tokenizer.vocab | Tokenizer | 10.4 KB | — | | vocab.txt | Tokenizer | 5.1 KB | — | ## License and Download License cc-by-4.0 Access Open weights, no gate Download size 2.5 GB [Download from MLX Community](https://huggingface.co/mlx-community/parakeet-tdt-0.6b-v2) Released by MLX Community through its official repository on Hugging Face. [Read the license](https://creativecommons.org/licenses/by/4.0/). ## Built From - Derived from nvidia/parakeet-tdt-0.6b-v2 ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 2.5 GB | | 16-bit | 1.2 GB | | 8-bit | 0.6 GB | | 4-bit | 0.3 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About parakeet-tdt-0.6b-v2 ### How much GPU memory does parakeet-tdt-0.6b-v2 need? About 1.5 GB at 16-bit and 0.4 GB at 4-bit: the weights (618M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run parakeet-tdt-0.6b-v2 on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use parakeet-tdt-0.6b-v2 commercially? Yes. parakeet-tdt-0.6b-v2 is released under Creative Commons Attribution 4.0. CC BY 4.0 permits sharing and adapting the work, including commercially, provided the creator is credited and changes are indicated. ## Similar Models Model · Speech recognition ### [parakeet-tdt-0.6b-v3](https://savrn.com/models/parakeet-tdt-0-6b-v3) [MLX Community](https://savrn.com/model-publishers/mlx-community) This model was converted to MLX format from nvidia/parakeet-tdt-0.6b-v3 using the conversion script. Please refer to original model card for more details on the model. Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/parakeet-tdt-0-6b-v3) Model · Speech recognition ### [fluid-speech-mini-mlx-4bit](https://savrn.com/models/fluid-speech-mini-mlx-4bit) [David Larrea](https://savrn.com/model-publishers/davidalarrea) Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/fluid-speech-mini-mlx-4bit) Model · Speech recognition ### [fluid-speech-mini-mlx-8bit](https://savrn.com/models/fluid-speech-mini-mlx-8bit) [David Larrea](https://savrn.com/model-publishers/davidalarrea) Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/fluid-speech-mini-mlx-8bit) Model · Speech recognition ### [underdog-bark-3.0-clean-mlx-4bit](https://savrn.com/models/underdog-bark-3-0-clean-mlx-4bit) [David Larrea](https://savrn.com/model-publishers/davidalarrea) Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/underdog-bark-3-0-clean-mlx-4bit) Model · Speech recognition ### [underdog-bark-3.0-clean-mlx-8bit](https://savrn.com/models/underdog-bark-3-0-clean-mlx-8bit) [David Larrea](https://savrn.com/model-publishers/davidalarrea) Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/underdog-bark-3-0-clean-mlx-8bit) Model · Speech recognition ### [underdog-bark-3.0-plus-mlx-4bit](https://savrn.com/models/underdog-bark-3-0-plus-mlx-4bit) [David Larrea](https://savrn.com/model-publishers/davidalarrea) Open weights cc-by-4.0 627M parameters mlx [View model](https://savrn.com/models/underdog-bark-3-0-plus-mlx-4bit) ## MLX Community [All models and datasets](https://savrn.com/model-publishers/mlx-community) ## Versions - [8ae155301e23](https://savrn.com/models/parakeet-tdt-0-6b-v2/versions/8ae155301e23) · current 2026-09-20 ## Explore More - [All speech recognition models](https://savrn.com/models/tasks/speech-recognition) - [All models under cc-by-4.0](https://savrn.com/models/licenses/cc-by-4-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [Hugging Face record](https://huggingface.co/mlx-community/parakeet-tdt-0.6b-v2) - [How the hub is built](https://savrn.com/model-hub/methodology)