Available today
Open weights to build on
Explore the published 8B, 30B and 123B models, their source foundations, licences and evaluation records. Try a bounded infrastructure-as-code task using the setup guidance in each model card.
Vinci Cyber by SimpleDirect® gives government and enterprise teams open-weight models to build their own defensive AI capability. Evaluate them on infrastructure you control, starting with human-reviewed infrastructure configuration work.
Start with a published release. Broader security workflows remain a development direction.

Vinci Cyber
A model can help prepare a change. Your organization needs the people, evidence and controls to decide whether it should happen. Vinci Cyber is a starting point for building that capability on infrastructure you choose.
Available today
Explore the published 8B, 30B and 123B models, their source foundations, licences and evaluation records. Try a bounded infrastructure-as-code task using the setup guidance in each model card.
Scoped engineering
Discuss a scoped project for private deployment, representative evaluations, workflow integration or model adaptation. Confirm feasibility, data rights, infrastructure, responsibilities and acceptance criteria before implementation.
Future direction
Our direction is to connect specialist models with permissioned security workflows, organization-specific knowledge and reviewable actions. Broader security-operations coverage must earn its place through evaluation; it is not established by today’s configuration-repair evidence.
Released weights are the starting point. A private deployment, integration or adapted model is a separately scoped engineering project. Future capabilities are goals, with availability and acceptance criteria still to be established.
Given a configuration and a scanner finding, Vinci Cyber is trained to propose a bounded repair, leave a configuration unchanged when appropriate, or reconsider an unsuccessful attempt.
The intended audience is security teams and engineers investigating infrastructure configuration findings. It is a narrow defensive specialization, not a general vulnerability scanner or an autonomous remediation system.
Illustrative evaluation workflow. These steps describe a trial to build, not a live product demonstration.
Choose a non-sensitive test configuration and a scanner finding from a system you are authorized to assess. Include the intended security and availability requirements.
Ask for a narrow repair, or a reason to leave the file unchanged. Compare the suggestion with the original file and check whether it introduces new findings or breaks expected behaviour.
Run independent checks and have an authorized engineer approve any change. Record the reasoning and outcome, including when the right answer is to stop.
SimpleDirect® specializes existing IBM Granite and Mistral Devstral foundations through post-training. These releases are not foundation models pretrained from scratch in Canada.
Choose a size to evaluate against your own configurations and acceptance criteria. Size alone is not a security-quality ranking; the 123B release uses a different foundation and licence.
Published release 1.0
A smaller starting point
The smallest weight footprint in this family. Consider it for an initial local workflow, then measure quality, latency and memory on your own task.
Explore Vinci Cyber 8B 1.0Published release 1.0
A larger Granite option
Compare this IBM Granite adaptation with 8B on the same cases. Plan for a larger weight footprint; a larger model does not guarantee a better result.
Explore Vinci Cyber 30B 1.0Published release 1.0
A different foundation
A Mistral Devstral adaptation with different infrastructure and licence considerations. Check deployment fit and the company-revenue condition before choosing it.
Explore Vinci Cyber 123B 1.0B indicates the approximate number of parameters in billions, not a quality score. Memory needs also depend on file format, context length, simultaneous requests and runtime; weight-file size is not a complete memory requirement.
Scroll the table horizontally for licences and downloads.
| Model | Foundation | Licence | Downloads |
|---|---|---|---|
| Vinci Cyber 8B 1.0 | ibm-granite/granite-4.1-8b | Apache-2.0 | Model card & weights (English) GGUF formats (English) |
| Vinci Cyber 30B 1.0 | ibm-granite/granite-4.1-30b | Apache-2.0 | Model card & weights (English) GGUF formats (English) |
| Vinci Cyber 123B 1.0 | mistralai/Devstral-2-123B-Instruct-2512 | Modified MIT (Mistral AI) | Model card & weights (English) GGUF formats (English) |
Licences and available formats are artifact-specific. Follow each model card for the detailed licence, loading examples and file checks.
The 8B and 30B releases use Apache-2.0. The 123B releases use Mistral’s Modified MIT licence, which includes a company-revenue condition. Review the linked terms before choosing a model for your organization.
Current model cards report a paired infrastructure-as-code evaluation across all three sizes and their own parent models. Results depend on the test set, runtime and format, and include failed repairs and unwanted changes. Scanner clearance alone does not establish a valid repair; syntax, security properties and operational impact still need independent checks.
Use each artifact’s own evaluation record. Results for a parent model, sibling size or full-precision checkpoint do not establish the performance of a different release or quantized format.
Public model cards provide weights and setup guidance for running Vinci Cyber on infrastructure you control. Independently check configuration syntax, security properties and regressions, then obtain human approval before applying a proposed change.
Allow memory beyond the weight files, choose a supported runtime, and record the exact revision and format. Tool use requires your own application, permissions and controlled workspace. Review and test suggested changes before applying them; these pages do not offer hosted inference.
For government and regulated work, your organization must assess the deployment and approve it for the information involved. Download availability does not establish security accreditation or suitability for a particular data classification.
Use the published weights independently under their licences. If you need engineering help, shape a project around your requirements.
Bring a concrete workflow and deployment requirements. Request implementation or adaptation directly, or scope a separate evaluation if suitability is still uncertain.
Set deliverables, acceptance checks, data rights, infrastructure ownership and procurement requirements. Confirm feasibility and responsibilities before implementation.
Decide who owns documentation, updates, monitoring and incident handling. Any ongoing support or operations are separately scoped, with availability and responsibilities agreed for the project.
Use the technical model cards alongside these guides and research records to decide what to test in your own environment.
Start with one useful workflow
Bring a non-confidential description of the work and ask for an implementation or adaptation proposal. A separate evaluation engagement is optional; deployment still needs agreed acceptance checks.
Opens your email app with an editable outline to [email protected]. Nothing is submitted by this page.
Keep the first message non-confidential. Do not send source code, security findings, documents, personal data or credentials. Agree on an appropriate sharing process first.