SAVRN
Search Contact SAVRN

Open-weight model · Any to any

Janus-Pro-1B

by DeepSeek-community deepseek-community/Janus-Pro-1B

Janus-Pro-1B is an open-weight model for any to any from DeepSeek-community, released under MIT License. It has 2.1B parameters and a 16,384-token context. At 16-bit it needs about 5 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 24.7k downloads a month.

Janus-Pro is a novel autoregressive framework that unifies multimodal understanding and generation.

Parameters2.1B
Context16,384
Weights4.2 GB
Licensemit
AccessOpen weights
Monthly Downloads24.7k

Runs On

What it takes to serve Janus-Pro-1B (2.1B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 4.2 GB 5.0 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 2.1 GB 2.5 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 1.0 GB 1.2 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x 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, read Oct 10, 2026.

Janus-Pro-1B on every accelerator the SAVRN Index prices, at every precision

Model Card

By DeepSeek-community, published under mit, revision 1655280bb759.

1. Introduction

Janus-Pro is a novel autoregressive framework that unifies multimodal understanding and generation. It addresses the limitations of previous approaches by decoupling visual encoding into separate pathways, while still utilizing a single, unified transformer architecture for processing. The decoupling not only alleviates the conflict between the visual encoder’s roles in understanding and generation, but also enhances the framework’s flexibility. Janus-Pro surpasses previous unified model and matches or exceeds the performance of task-specific models. The simplicity, high flexibility, and effectiveness of Janus-Pro make it a strong candidate for next-generation unified multimodal models.

Github Repository

2. Model Summary

Janus-Pro is a unified understanding and generation MLLM, which decouples visual encoding for multimodal understanding and generation. Janus-Pro is constructed based on the DeepSeek-LLM-1.5b-base/DeepSeek-LLM-7b-base.

For multimodal understanding, it uses the SigLIP-L as the vision encoder, which supports 384 x 384 image input. For image generation, Janus-Pro uses the tokenizer from here with a downsample rate of 16.

3. Usage Examples

Single Image Inference

Read the full model card (460 words)

Configuration

Architecture
JanusForConditionalGeneration
Context length (tokens)
16,384
Layers
24
Hidden size
2,048
Feed-forward size
5,632
Attention heads
16
Key/value heads
16
Head dimension
128
Vocabulary size
102,400
RoPE base
10000
Stored precision
bfloat16
Model type
janus

Identity and Version

Repository
deepseek-community/Janus-Pro-1B
Publisher
DeepSeek-community
Task
Any to any
Modality
Multimodal
Library
transformers
Parameters
2.1B parameters
Languages
Not stated by the source
Revision
1655280bb75959cc1cb85529a2a8b26e7016072e
First published
2025-03-01
Last updated
2026-03-18

Files and Weights

13 files, 4.2 GB in total. The weights are 1 file totalling 4.2 GB in safetensors.

Weights1 file · 4.2 GB
Configuration5 files · 3.5 KB
Tokenizer2 files · 7.7 MB
Documentation1 file · 5.0 KB
Other3 files · 629.1 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model.safetensorsWeights4.2 GB 9d1a416f95fb
config.jsonConfiguration1.9 KB —
generation_config.jsonConfiguration233 B —
preprocessor_config.jsonConfiguration465 B —
processor_config.jsonConfiguration79 B —
special_tokens_map.jsonConfiguration792 B —
README.mdDocumentation5.0 KB —
chat_template.jinjaOther1.0 KB —
janus_pro_teaser1.pngOther98.0 KB —
janus_pro_teaser2.pngOther530.0 KB dcd9be432b9d
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer7.6 MB —
tokenizer_config.jsonTokenizer105.9 KB —

License and Download

License
mit
Access
Open weights, no gate
Download size
4.2 GB
Download from DeepSeek-community

Released by DeepSeek-community through its official repository on Hugging Face. Read the license.

Built From

  • Described by arXiv:2501.17811

Memory Requirements

PrecisionWeights in memory
As published4.2 GB
16-bit4.2 GB
8-bit2.1 GB
4-bit1.0 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About Janus-Pro-1B

How much GPU memory does Janus-Pro-1B need?

About 5 GB at 16-bit and 1.2 GB at 4-bit: the weights (2.1B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run Janus-Pro-1B 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 Janus-Pro-1B commercially?

Yes. Janus-Pro-1B is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.

What is Janus-Pro-1B's context length?

16,384 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Any to any

gemma-4-31B-it-assistant

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 470M parameters 262,144 tokens transformers
View model

Model · Any to any

gemma-4-26B-A4B-it-assistant

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 420M parameters 262,144 tokens transformers
View model

Model · Any to any

gemma-4-E4B-it-assistant

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 79M parameters 131,072 tokens transformers
View model

Model · Any to any

gemma-4-12B-it-qat-GGUF

Unsloth AI

This model ships a Multi-Token Prediction drafter at the repo root (mtp-gemma-4-12B-it.gguf, a near-lossless smart Q40). A recent llama.cpp auto-discovers it from -hf, so you do not pass --model-draft: The drafter shares the target's KV cache and does not change the output (the target verifies every drafted token). See the MTP/ folder for the other precisions and explicit usage. Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context…

Open weights apache-2.0 transformers
View model

Model · Any to any

gemma-4-E4B-it-GGUF

GGML Org

Run with https://llama.app - https://huggingface.co/google/gemma-4-E4B-it - https://huggingface.co/google/gemma-4-E4B-it-assistant - https://huggingface.co/google/gemma-4-E4B-it-qat-q40-unquantized-assistant - https://huggingface.co/google/gemma-4-E4B-it-qat-q40-unquantized - add info - add dflash

Open weights apache-2.0
View model

Model · Any to any

gemma-4-12B-it-qat-q4_0-gguf

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 transformers
View model