◆ Gentoo Logic · Modeling Warehouse
Warehouse › AI-coding › Magic

AI-coding · Multi-Cloud

Magic

Builds its own ultra-long-context (LTM) code models for autonomous software engineering.

Foundation proximity: Native (own / self-hosted weights)

Foundational models & how they consume them

PrimaryMagic LTM (own)· self-hosted · Native (own / self-hosted weights)

Cloud & datastore

Multi-Cloudobject

Compliance

—

Similar companies

PoolsideTabnineCognition (Devin)LovableWindsurf (Codeium)
Head-to-head: Magic vs PoolsideMagic vs TabnineMagic vs Cognition (Devin)

Others that build on proprietary / self-built models

Abnormal SecurityAbridgeAdobeAI21 LabsAleph AlphaAmbience

Explore

More AI-coding →Native (own / self-hosted weights) →More on Multi-Cloud →

FAQ

Does Magic use a foundational AI model?
Magic uses Magic LTM (own) (Native (own / self-hosted weights)).
What cloud does Magic run on?
Magic runs on Multi-Cloud.
What database does Magic use?
Magic uses object.
How close does Magic run to the foundational model?
Magic is Native (own / self-hosted weights) — close to the metal, so cheaper tokens and more control over outputs.

Sources

magic.dev ↗

Also on the map

proprietary / self-built models

← Full orbit map (260 companies) · Score your own stack →

Architecture inferred from public sources · confidence low · verify before betting on a detail.