SAVRN
Search Contact SAVRN

Open-weight model · Image and text to text

RICO

by Sonny DeSorbo darkc0de/RICO

RICO is an open-weight model for image and text to text from Sonny DeSorbo. It has 27.8B parameters and a 262,144-token context. At 16-bit it needs about 66.7 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 429 downloads a month.

This model is a part of The XORTRON Criminal Computing project; an ongoing research experiment and exercise in AI safety and alignment.

Parameters27.8B
Context262,144
Weights55.6 GB
License—
AccessOpen weights
Monthly Downloads429

Runs On

What it takes to serve RICO (27.8B 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 55.6 GB 66.7 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 27.8 GB 33.3 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 13.9 GB 16.7 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 Sep 25, 2026.

RICO on every accelerator the SAVRN Index prices, at every precision

Model Card

This model is a part of The XORTRON Criminal Computing project; an ongoing research experiment and exercise in AI safety and alignment. XORTRON models are intentionally developed to study, evaluate, and document the potential of advanced AI systems to facilitate real-world criminal activity, abuse, and other high-risk conduct. Because these models may produce unusually capable, operationally relevant, or otherwise sensitive outputs, they are not intended for unrestricted public use. By requesting access to, downloading, cloning, copying, deploying, or otherwise obtaining this model or its weights, you certify that you are accessing it in a legitimate professional, governmental, legal…

Excerpt from the card by Sonny DeSorbo.

Configuration

Architecture
Qwen3_5ForConditionalGeneration
Context length (tokens)
262,144
Layers
64
Hidden size
5,120
Feed-forward size
17,408
Attention heads
24
Key/value heads
4
Head dimension
256
Vocabulary size
248,320
Stored precision
bfloat16
Model type
qwen3_5

Identity and Version

Repository
darkc0de/RICO
Publisher
Sonny DeSorbo
Task
Image and text to text
Modality
Image and text
Library
transformers
Parameters
27.8B parameters
Languages
en
Revision
ca3c0d053313aeb2f28c47a2fbd4638294240c25
First published
2026-09-07
Last updated
2026-09-25

Files and Weights

27 files, 55.6 GB in total. The weights are 18 files totalling 55.6 GB in safetensors.

Weights18 files · 55.6 GB
Configuration4 files · 118.0 KB
Tokenizer2 files · 20.0 MB
Documentation1 file · 8.7 KB
Other1 file · 28.2 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00018.safetensorsWeights4.0 GB 0c4a3394914c
model-00002-of-00018.safetensorsWeights3.0 GB 7d297fcace9d
model-00003-of-00018.safetensorsWeights2.5 GB e6fc252b2278
model-00004-of-00018.safetensorsWeights4.0 GB bb9990af635c
model-00005-of-00018.safetensorsWeights2.1 GB 27199a636af5
model-00006-of-00018.safetensorsWeights4.0 GB 1c1ec960210c
model-00007-of-00018.safetensorsWeights2.1 GB 872199bf1d28
model-00008-of-00018.safetensorsWeights4.0 GB 05da60209559
model-00009-of-00018.safetensorsWeights2.1 GB 9eeed4bce052
model-00010-of-00018.safetensorsWeights4.0 GB 106067bb3189
model-00011-of-00018.safetensorsWeights2.1 GB 8a74338e09d9
model-00012-of-00018.safetensorsWeights4.0 GB c92cdfd69a11
model-00013-of-00018.safetensorsWeights2.1 GB 7b272602b154
model-00014-of-00018.safetensorsWeights4.0 GB e923230a1a7d
model-00015-of-00018.safetensorsWeights2.1 GB 4d84fc191210
model-00016-of-00018.safetensorsWeights4.0 GB 939540187e29
model-00017-of-00018.safetensorsWeights2.1 GB fc24da59137f
model-00018-of-00018.safetensorsWeights3.4 GB 1ac226c56138
config.jsonConfiguration4.3 KB —
generation_config.jsonConfiguration213 B —
model.safetensors.index.jsonConfiguration112.2 KB —
processor_config.jsonConfiguration1.3 KB —
README.mdDocumentation8.7 KB —
chat_template.jinjaOther28.2 KB —
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer20.0 MB 87a7830d63fc
tokenizer_config.jsonTokenizer36.1 KB —

License and Download

License
Not stated by the source
Access
Open weights, no gate
Download size
55.6 GB
Download from Sonny DeSorbo

Released by Sonny DeSorbo through its official repository on Hugging Face.

Built From

  • Derived from Qwen/Qwen3.8-27B
  • Trained on (disclosed) darkc0de/XORTRON-RESTRICTED-RESEARCH-SFT

Memory Requirements

PrecisionWeights in memory
As published55.6 GB
16-bit55.6 GB
8-bit27.8 GB
4-bit13.9 GB

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

Built on This Model

Questions About RICO

How much GPU memory does RICO need?

About 66.7 GB at 16-bit and 16.7 GB at 4-bit: the weights (27.8B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run RICO 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.

What is RICO's context length?

262,144 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Image and text to text

Qwen3.8-27B

Qwen

Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be…

Open weights apache-2.0 27.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.8-27B-FP8

Qwen

Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be…

Open weights apache-2.0 27.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.6-27B

Qwen

Following the February release of the Qwen3.5 series, we're pleased to share the first open-weight variant of Qwen3.6. Built on direct feedback from the community, Qwen3.6 prioritizes stability and real-world utility, offering developers a more intuitive, responsive, and genuinely productive coding experience. This release delivers substantial upgrades, particularly in For more details, please refer to our blog post Qwen3.6-27B. Empty cells (--) indicate scores not yet available or not applicable. For streamlined integration, we recommend using Qwen3.6 via APIs. Below is a guide to use Qwen3.6 via OpenAI-compatible API. Qwen3.6 can be served via APIs with popular inference frameworks. In…

Open weights apache-2.0 27.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.5-27B

Qwen

Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to empower developers and enterprises with unprecedented capability and efficiency. For more details, please refer to our blog post Qwen3.5. WMT24++: a harder subset of WMT24 after difficulty labeling and rebalancing; we report the averaged scores on 55 languages using XCOMET-XXL. Empty cells (--) indicate scores not yet available or not applicable. Empty cells (--) indicate scores not…

Open weights apache-2.0 27.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

Qwen3.8-27B-AWQ-INT4

Cyankiwi

Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be…

Open weights apache-2.0 27.8B parameters 262,144 tokens transformers
View model

Model · Image and text to text

XORTRON-RICO-v3

Sonny DeSorbo

To achieve optimal performance, use the following settings: Thinking: temperature=1.0, topp=0.95, topk=20, minp=0.0, presencepenalty=0.0, repetitionpenalty=1.0 Instruct: temperature=0.7, topp=0.80, topk=20, minp=0.0, presencepenalty=1.5, repetitionpenalty=1.0 This model is a part of The XORTRON Criminal Computing project; an ongoing research experiment and exercise in AI safety and alignment. XORTRON models are intentionally developed to study, evaluate, and document the potential of advanced AI systems to facilitate real-world criminal activity, abuse, and other high-risk conduct. Because these models may produce unusually capable, operationally relevant, or otherwise sensitive outputs…

Open weights 27.8B parameters 262,144 tokens transformers
View model