AI Privacy,End to End

Real private AI means the model runs inside GPU TEE, model hosting emits verifiable privacy proofs, and the gateway is verifiable too.

GPU TEE Models

Inference runs inside NVIDIA Confidential Compute.

Hosting Proofs

Users can verify model-hosting privacy integrity.

Gateway Proof

RedPill gateway routing and receipts are verifiable.

Get API Key Try the Chat

Four Reasons to Use RedPill

RedPill aggregates private AI supply behind one gateway: more model choice, better prices, TEE-backed security, and one verification workflow for every provider.

Most Private Models

Access private GPU models across Phala, Tinfoil, NEAR AI, Chutes, and new verified providers without rebuilding your app for each network.

TEE-Secured Gateway

Requests pass through a gateway running in a trusted execution environment, with encrypted paths and key binding for sensitive prompts and responses.

Better Pricing

Compare unified token prices and route across providers by cost, speed, and availability without exposing provider complexity to your application.

Unified Verification

One evidence workflow covers gateway attestation, provider verification, signed receipts, and audit sessions across every supported private model provider.

Up and running in minutes

YOUR SDK.
PRIVATE AI.

Keep the OpenAI client you already use. Change the base URL, choose a private model, and add verification when your workflow needs attestation reports, response receipts, and audit sessions.

View Full Docs Private AI Gateway RedPill GitHub

chat.py

import osfrom openai import OpenAIclient = OpenAI(    base_url="https://api.redpill.ai/v1",    api_key=os.environ["REDPILL_API_KEY"],)response = client.chat.completions.create(    model="openai/gpt-oss-120b",    messages=[        {"role": "user", "content": "Hello, confidential world"}    ],    extra_body={"provider": {"aci_verified": True}},)print(response.id)print(response.choices[0].message.content)

Built for How You Work

Build private AI products, protect regulated workflows, and keep encrypted chat available as a proof-of-product.

One Gateway, Private Models

Build privacy-first applications without juggling provider credentials. Switch in minutes: just change the base URL.

Full API surface: chat, embeddings, audio, images, files, batches, and models
Private routing: never downgrades from a protected backend to a standard fallback
Virtual keys: per-key budgets, RPM/TPM limits, and routing strategy

Get API Key

Ship Faster, Stay Compliant

Building healthcare, legal, or fintech apps? RedPill handles private inference, verification, and encrypted routing so you can focus on your product.

Attestation APIs for nonce-bound hardware reports
Open source gateway and verifier
Verifier CLI for independent proof checks

View Documentation

Private AI Model Catalog

The most private model options in one place, with unified pricing, provider coverage, and live performance through one gateway.

Z.ai: GLM 5.2

Private GPU

GLM-5.2 is Z.ai's flagship model for the era of long-horizon tasks. With a truly usable 1M-token context window, it can handle project-level engineering context and execute long-running tasks more reliably. Served as a text-only TEE deployment via Phala.

by Phala|1M context|$1.40/M input|$4.40/M output|$0.26/M cache read

Intel TDXNVIDIA CC

Google: Gemma 4 31B

Private GPU

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense model. Features a 256K token context window, configurable thinking/reasoning mode, native function calling, and strong multilingual performance. Served as a text-only TEE deployment via NEAR AI.

by Phala|262K context|$0.15/M input|$0.46/M output|$0.075/M cache read

Intel TDXNVIDIA CC

Phala: Gemma-4 26B-A4B Uncensored (Heretic)

Private GPU

Uncensored "Heretic" variant of google/gemma-4-26B-A4B-it created using Heretic v1.2.0 with the Arbitrary-Rank Ablation (ARA) method and row-norm preservation. Refusals drop from 100/100 to 11/100 with KL divergence 0.0499 vs the base model. The base Gemma 4 26B A4B is a Mixture-of-Experts model with 25.2B total / 3.8B active parameters (8 active / 128 total experts), 30-layer transformer with hybrid local sliding (1024) + global attention, supporting a 256K context window. Natively multimodal (text + images, variable aspect ratios). Strong on coding, reasoning, function calling, with native system prompt support across 35+ languages. Served on Phala in TDX-attested H200 enclave with end-to-end ECDSA response signing; vLLM-compatible FP8-Static quantization by cloud19 (router excluded from quantization).

by Phala|66K context|$0.15/M input|$0.70/M output

Intel TDXNVIDIA CC

Phala: Qwen3.6 35B-A3B Uncensored (Aggressive)

Private GPU

Uncensored "Aggressive" variant of Qwen3.6-35B-A3B from Alibaba's Qwen team. The fine-tune by HauhauCS removes refusal behaviors (0/465 refusals) without modifying datasets or core capabilities. The base architecture is a 35B-parameter Mixture-of-Experts model with 256 experts routing 8 per token (~3B active params), 40 layers, and a hybrid linear+full-softmax attention mechanism (3:1 ratio). Supports a native 262K context and is natively multimodal across text, images, and video. Served on Phala in TDX-attested H200 enclave with end-to-end ECDSA response signing; FP8 quantization by lamianlbe.

by Phala|131K context|$0.30/M input|$1.50/M output

Intel TDXNVIDIA CC

Sentence Transformers: all-MiniLM-L6-v2

Private GPU

The all-MiniLM-L6-v2 embedding model maps sentences and short paragraphs into a 384-dimensional dense vector space, enabling high-quality semantic representations that are ideal for downstream tasks such as information retrieval, clustering, similarity scoring, and text ranking.

by Phala|512 context|$0.005/M input|$0.00/M output

Intel TDXNVIDIA CC

Qwen2.5 7B Instruct

Private GPU

Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:

Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.
Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.
Long-context Support up to 128K tokens and can generate up to 8K tokens.
Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.

Usage of this model is subject to .

by Phala|33K context|$0.10/M input|$0.20/M output

Intel TDXNVIDIA CC

Browse Private Models

Private AI Models

No memory. No traces. The model knows nothing about you.

Privacy Off

Standard AI API

01Provider can inspect prompts

Logs depend on policy controls

03No per-response proof trail

Prompts, route details, and audit evidence are scattered across provider systems.

Privacy On

RedPill Gateway

01Encrypted prompt and response paths

Verified private provider route

03Receipt and audit session available

One encrypted gateway returns route, receipt, and verification evidence.

Normal AI API vs Private AI Gateway

Privacy policies ask for trust. RedPill gives developers encrypted requests, verified private providers, signed receipts, and a single workflow for checking the evidence.

OpenAIAnthropicGooglexAIOracle

Normal AI API

Privacy depends on policy language
Provider proof is fragmented or unavailable
No unified verification workflow
No per-response receipt or audit session

PhalaTinfoilNEAR AIChutes

RedPill Private AI

Verified private model routing
End-to-end encrypted AI paths
Live price, speed, and uptime routing
Signed receipts and verifier tooling

Verify It

Trusted by Industry Leaders

From AI infrastructure providers to law firms and security companies. RedPill powers privacy-first AI for organizations that can't compromise on confidentiality.

Try RedPill Free

Our Partners & Integrations

Cameron, Director of Near AI

"RedPill's confidential computing approach aligns perfectly with our vision for decentralized AI. Their TEE infrastructure sets a new standard for privacy-preserving inference at scale."

Elizabeth Leon Gonzalez, Milligan, Beswick, Levine & Knox LLP

"Attorney-client privilege is non-negotiable. RedPill is the only AI platform our firm trusts for case research and contract analysis. The cryptographic guarantees give us confidence no other tool can."

Vin Sharma, Founder at Vijil AI

"As a security-focused AI company, we evaluated every private AI solution on the market. RedPill's end-to-end encryption and TEE architecture passed our most rigorous security audits."

Chris Were, CEO at Blue Nexus AI

"We built Blue Nexus on the belief that personal AI should be truly personal. RedPill's infrastructure lets us deliver that promise: your AI assistant that never shares your secrets."

Start with Private AI

Developer-first plans for encrypted AI, enterprises, and private chat users.

API First

Developer

Build with private GPU models, smart routing, and a drop-in API. Pay with card or crypto.

Pay as you go

200+ models through one endpoint
Private GPU models across 4 networks
Virtual keys with budget and rate limits
Verification reports, receipts, and audit sessions
Stripe and Coinbase Commerce payments

Get API Key

Enterprise

For teams needing SSO, admin controls, audit workflows, and dedicated private inference support.

Custom Pricing

Everything in Developer
Enterprise virtual-key policies
Compliance support and audit evidence
Dedicated support and invoicing
Deployment architecture review

Talk to Sales

Pro

For professionals who want a private chat experience backed by encrypted AI infrastructure.

$35/month (billed annually)

Private chat access
Broad model catalog
Encrypted conversation workflows
File upload in the chat app
Verification UX in product

Get Pro

Free

Try the chat product and see confidential model access before adding payment.

Get Started

Private chat access
Limited daily usage
E2EE-supported workflows
Web access

Try the Chat

Ready to Build AI People Trust?

Schedule a demo to see how RedPill can secure your AI use cases.

Schedule a DEMO View Pricing

Frequently Asked Questions

Practical details for building with end-to-end encrypted AI

Enterprise Ready

Complete Compliance & Security Readiness

RedPill meets the highest security and compliance standards for regulated industries. Our TEE-based architecture ensures your data stays protected while meeting enterprise requirements.

SOC 2 Type II Certified

Annual audits verify our security controls for data protection, availability, and confidentiality meet enterprise standards.

HIPAA Compliant

Healthcare organizations can safely use RedPill with protected health information. BAAs available for enterprise customers.

GDPR & ISO 27001

Full compliance with EU data protection regulations and internationally recognized information security management.

Get an API key in 2 minutes

Start with the OpenAI SDK, choose a private model, and add verifier checks when you need audit-grade proof.

Get API Key

Private Chat

Private model

Hi. Ask me anything and I'll open the confidential chat with openai/gpt-oss-120b selected.

Zero data retentionVerifier ready

End-to-end encrypted AI through one gateway: private models, verified providers, unified pricing, live performance, and evidence your team can check.

AI Privacy,End to End