Universal-1

Closed weights AssemblyAI 600M parameters April 2024

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
AssemblyAI
Organisation type
Industry
Published
3 April 2024
Authors
Francis McCann (lead), Luka Chkhetiani (lead), Andrew Ehrenberg, Robert McHardy, Rami Botros, Yash Khare, Andrea Vanzo, Taufiquzzaman Peyash, Gabriel Oexle, Michael Liang, Ilya Sklyar, Ahmed Etefy, Daniel McCrystal, William Pipsico Ferreira, Ruben Bousbib, Ben Gotthold, Soheyl Bahadoori, Enver Fakhan, Rahul Bagai, Mez Rashid, James He, Takuya Yoshioka, Travis Kupsche, Domenic Donato, Marco Ramponi…

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Speech
Task
Speech recognition (ASR)
Base model
Conformer

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
600M

"Training on more than 12.5 million hours of diverse multilingual audio data, 600M-parameter Conformer RNN-T based Universal-1 achieves remarkable robustness"

Training data
tokens

"Trained on over 12.5 million hours of multilingual audio data" + fine-tuning 150K hours "A batch size of 2048 is used."

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
Google TPU v5e
Hardware utilisation
MFU 54.0%

"Pre-training was performed on a cluster of v5e TPU chips, operating with 54% Model Flops Utilization (MFU), thanks to our optimized JAX implementation."

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
API access
Training code
Unreleased

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Record confidence
Confident

Sources

Where this record came from and when it was last checked.

Reference
Robust and accurate multilingual speech-to-text
Last updated
28 November 2025

What the numbers mean

Where it came from

Universal-1 was published by AssemblyAI, during April 2024. The category the publisher falls under is industry.

It works in the domain of Speech, and is recorded as performing the task of speech recognition (ASR).

Its starting point was an existing base model, Conformer. Most models at this scale are adapted from an existing base rather than built from nothing.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

Answers

Universal-1 — common questions

01

Universal-1— what GPU do I need to run it?

None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

02

Universal-1— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

03

Universal-1— how many parameters does it have?

It has a parameter count of 600M. "Training on more than 12.5 million hours of diverse multilingual audio data, 600M-parameter Conformer RNN-T based Universal-1 achieves remarkable robustness". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

04

Universal-1— who created it?

It was published by AssemblyAI, an organisation categorised as industry.

05

Universal-1— when was it released?

It was published in April 2024. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

06

Universal-1— what is it used for?

It works in the domain of Speech, and is recorded as handling the task of speech recognition (ASR). Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Source

Original publication

Record last updated 28 November 2025

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.