Models Overview

ORGN Gateway exposes models through two distinct execution environments, each with different privacy guarantees: Trusted Execution Environments (TEE) and Zero Data Retention (ZDR). Both ensure your inference data is not stored or logged, but they differ in how that guarantee is enforced and what evidence you receive.

Execution types

Trusted Execution Environment (TEE)

TEE models run inside hardware-isolated Trust Domains. The CPU and GPU execute inference in an environment that is encrypted and isolated from the host operating system, the hypervisor, and infrastructure personnel — including Gateway operators and the model provider.

Every TEE inference request produces a cryptographic attestation receipt: hardware-signed evidence that proves which model ran, inside which verified environment, and that execution was not tampered with. This is hardware-enforced privacy you can independently verify in Scanner.

TEE models run on infrastructure provided by NEAR and Phala Network, both operating Intel TDX–based confidential virtual machines with NVIDIA H100 GPU attestation.

What TEE guarantees:

• Prompts and responses encrypted in memory during execution, invisible to host OS, hypervisor, cloud provider, and Gateway
• Hardware-signed attestation receipt per request, verifiable against Intel and NVIDIA public PKI
• Cryptographic proof that the exact model you requested ran inside a genuine, unmodified Trust Domain
• Zero content retention: no prompts or outputs stored or logged

TEE infrastructure providers:

Zero Data Retention (ZDR)

ZDR models run on Vercel's AI infrastructure under contractual zero data retention commitments from underlying model providers. Vercel's AI gateway enforces that providers do not store, log, or use your prompts and responses for any purpose, including model training.

ZDR does not use hardware-isolated execution environments. There is no attestation receipt and no cryptographic proof of execution. The privacy guarantee is policy-enforced, not hardware-verified.

ZDR opens a dramatically larger catalog: frontier models from Anthropic, OpenAI, Google, Meta, Mistral, and dozens more — plus image generation, video generation, and embedding models not available in TEE environments.

What ZDR guarantees:

• Inference providers do not store or log your prompts or outputs
• No training on your data
• Policy-enforced zero retention by Vercel and underlying providers
• Broadest frontier model catalog under one API key

ZDR infrastructure provider:

Comparison

Choosing between TEE and ZDR

Choose TEE when:

• You operate in a regulated industry and need hardware-level data isolation
• You need cryptographic proof of execution for audit or compliance
• Your threat model includes infrastructure-level compromise or insider risk at the provider
• You require independently verifiable privacy guarantees per request

Choose ZDR when:

• You need frontier closed-weight models (Claude, GPT, Gemini) not yet in TEE
• Your use case requires image generation, video generation, or advanced embedding models
• Policy-enforced zero retention satisfies your compliance requirements
• You want the broadest model catalog under a single API key

Both model types use the same OpenAI-compatible Gateway API at https://api.gateway.orgn.com/v1. Model IDs use underscores (near_glm_4_7, vercel_claude_sonnet_4_6). The model ID you send determines the execution environment.

Browse by modality