Skip to content
@Aevonix

Aevonix Research

Open source AI research. HPC equipment sales and operations.
Aevonix Research

Open source AI research. HPC equipment sales and operations.

We develop local agentic systems, incubate AI projects and support the pursuit of open access to AGI.

Research

Protagine

Our work toward Proto-AGI is open source and ongoing. Protagine (Proto AGI Engine) is a mind for Hermes agents: persistent, source-backed memory across conversations, channels and model changes; commitments it keeps; drives that notice what matters and take initiative within the owner's authority; outreach that learns from the owner's replies; and typed decisions with an optional fast local decision model. It attaches to stock Hermes, runs on local inference, and measures every capability against stock Hermes by paired ablation. V1 of the Proto-AGI line is out, with its evidence and its open gaps in the repository.

Protagine logo downloads · Brand guide

Open recipes. MiMo V2.6 Pro RL on eight DGX Sparks: vLLM patches, a DFlash recipe and reproducible results.

HPC equipment and operations

We work with OEMs and resellers to help customers gain priority access to HPC equipment we sell. We also support datacenter placement, cluster management and monetization through our partners and network.

We have supported several NVIDIA B300 deployments across Central America and the United States.

HPC investment SPVs

We are launching HPC investment SPVs in 2027.

Website · Branding & press kit · contact@aevonix.com

Popular repositories Loading

  1. GLM-5.3-Flash-EXL3-4x-RTX-PRO-6000-TensorFold GLM-5.3-Flash-EXL3-4x-RTX-PRO-6000-TensorFold Public

    GLM-5.3-Flash EXL3 on 4x RTX PRO 6000 with TensorFold: one-command recipe and engine patches. By Aevonix Research in collaboration with Mia's AI Lab.

    Python 5 1

  2. mimo-2.6-dgx-spark mimo-2.6-dgx-spark Public

    MiMo V2.6 Pro RL on eight DGX Sparks: vLLM patches, DFlash recipe, and reproducible results.

    Python 4

  3. .github .github Public

    Aevonix Research organization profile and brand assets.

Repositories

Showing 3 of 3 repositories
  • GLM-5.3-Flash-EXL3-4x-RTX-PRO-6000-TensorFold Public

    GLM-5.3-Flash EXL3 on 4x RTX PRO 6000 with TensorFold: one-command recipe and engine patches. By Aevonix Research in collaboration with Mia's AI Lab.

    Aevonix/GLM-5.3-Flash-EXL3-4x-RTX-PRO-6000-TensorFold's past year of commit activity
    Python 5 Apache-2.0 1 0 0 Updated Oct 3, 2026
  • .github Public

    Aevonix Research organization profile and brand assets.

    Aevonix/.github's past year of commit activity
    0 0 0 0 Updated Sep 27, 2026
  • mimo-2.6-dgx-spark Public

    MiMo V2.6 Pro RL on eight DGX Sparks: vLLM patches, DFlash recipe, and reproducible results.

    Aevonix/mimo-2.6-dgx-spark's past year of commit activity
    Python 4 Apache-2.0 0 0 0 Updated Sep 23, 2026

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…

Most used topics

Loading…