◆ Gentoo Logic · Modeling Warehouse

Who builds on Meta · Llama

15 companies build on Meta · Llama

15 companies in the Modeling Warehouse build on Meta · Llama, ranked by how directly they consume it — Native first, Middleware last. Avg Foundation Proximity Score 97/100.

Native 14Direct API 0Cloud-hosted 1Middleware 0

Baseten

AI-infra · Multi-Cloud · Native (own / self-hosted weights)

Model-inference platform serving open models (Llama, etc.) on autoscaling GPU infra.

Cerebras

AI-hardware · On-Prem · Native (own / self-hosted weights)

Designs wafer-scale AI chips and serves open models (Llama) at record speed on its own systems.

Cloudflare

Infrastructure · On-Prem · Native (own / self-hosted weights)

Operates its own global edge network of data centers (not a hyperscaler tenant). Developer platform: Workers (V8 isolates), D1/KV/R2 storage, and Workers AI serving open models at the edge.

Databricks

Data Platform · Multi-Cloud · Native (own / self-hosted weights)

Lakehouse deployed into the customer's own cloud (AWS/Azure/GCP), built on open Spark + Delta Lake. Ships its own open model DBRX and the Mosaic AI training/serving stack.

Dell

Hardware/Enterprise · Hybrid · Native (own / self-hosted weights)

Dell AI Factory packages on-prem open models (Llama) + partners (Cohere) for enterprise/sovereign deployment.

Fireworks AI

AI Infra · Multi-Cloud · Native (own / self-hosted weights)

Fast inference for open models on its own serving stack.

Groq

AI Hardware · On-Prem · Native (own / self-hosted weights)

Designs its own LPU inference chips; serves Llama + open models at extreme speed.

Hugging Face

AI/LLM · AWS · Native (own / self-hosted weights)

The model/dataset hub on AWS, storing weights as Git LFS repos over S3 and running Inference Endpoints/Spaces. Stewards open libraries (Transformers, Diffusers) — the open-source center of gravity for ML.

Lightning AI

AI Infra · Multi-Cloud · Native (own / self-hosted weights)

Builds PyTorch Lightning + a studio to train/serve open models (Llama, etc.) on cloud GPUs.

Replicate

AI Infra · GCP · Native (own / self-hosted weights)

Runs open models as one-click APIs on its own GPU fleet.

SambaNova

AI-hardware · On-Prem · Native (own / self-hosted weights)

Own RDU AI chips plus its Samba-1 model and hosted open models for enterprise/government.

Scale AI

AI/Data · AWS · Native (own / self-hosted weights)

Beyond data-labeling, Scale trains its own Defense Llama (fine-tuned Llama 3) for U.S. national security on its Donovan federal platform; evaluates frontier models for the DoD CDAO.

Snowflake

Data Platform · Multi-Cloud · Native (own / self-hosted weights)

Cloud data warehouse that runs as a tenant across AWS, Azure and GCP, separating compute from storage on object stores; metadata in FoundationDB. Cortex serves hosted LLMs next to the data.

Together AI

AI Infra · Multi-Cloud · Native (own / self-hosted weights)

Serves open models (Llama, Mistral, etc.) on its own GPU cloud — the infrastructure to run Native.

Oracle

Enterprise · On-Prem · Cloud-hosted (Bedrock/Vertex/Azure)

Runs its own OCI; OCI Generative AI hosts Cohere + Llama for enterprise.

← Back to the full orbit map (all 260 companies) · Score your own stack →