Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Privatemode AI logo

Provider details

Privatemode AI

Learn more about this provider, then browse the models currently listed under it.

privatemode-aiprivatemode-ai

Latest news about Privatemode AI

Privatemode AI

Official sourceAnnouncement

Edgeless Systems announced on March 12, 2026 that Privatemode AI brings confidential AI to government, healthcare, and other regulated industries across Europe. The first-party press index says the service is built on NVIDIA Confidential Computing and provides cryptographically verifiable AI inference designed to prote The same Edgeless Systems press page also dates the original Privatemode AI launch to February 19, 2025, describing it as an always-encrypted AI service offered through both a desktop app and an API. The supplied excerpt does not provide model versions, API changes, pricing, benchmarks, or other implementation details.

Privatemode AI

Official sourceRelease Notes

Privatemode's official release notes document a series of provider-level updates spanning versions v1.44.0 through v1.54.0, with a hard cutoff notice that "versions before v1.46.0 don't work anymore since August 2, 2026." The most recent entry, v1.54.0, adds a new GLM-5.2 model preview accessible via the model IDs glm- Earlier entries ship additional developer-facing changes: v1.53.0 introduces a JavaScript/TypeScript SDK for Privatemode; v1.52.0 upgrades vLLM to v0.27.1 and Contrast to v1.23.0; v1.51.0 adds a DeepSeek OCR 2 model preview, retires Gemma 4 31B in favor of DeepSeek OCR 2 or Kimi K2.6 for vision, and introduces a shared

About Privatemode AI

What they do

Privatemode AI provides confidential AI endpoints with end-to-end encryption and OpenAI- and Anthropic-compatible APIs for chat, embeddings, and speech workloads.

How they were founded

Privatemode AI launched in February 2025 from Edgeless Systems, a confidential-computing company founded by Felix Schuster and Thomas Tendyck.

Quick Info

Organization
Edgeless Systems
Headquarters
Bochum, North Rhine-Westphalia, Germany
SDK package
@ai-sdk/openai-compatible
Synced at
Jul 11, 2026

Models served

11 models available through Privatemode AI

These model results are sorted newest first so you can quickly see the latest options from this provider.

Privatemode AIglm-flashbeta
GLM Flash (latest)
glm-flash-latestPrivatemode AIReleased Aug 26, 20261,000,000 token context windowIn $0.2311 · Out $0.7511
Input
Output
Privatemode AIglm-flashbeta
GLM-5.3-Flash
glm-5.3-flashPrivatemode AIReleased Aug 26, 20261,000,000 token context windowIn $0.2311 · Out $0.7511
Input
Output
Privatemode AIglmbeta
GLM (latest)
glm-latestPrivatemode AIReleased Aug 14, 20261,000,000 token context windowIn $1.791 · Out $6.6326
Input
Output
Privatemode AIglmbeta
GLM-5.3
glm-5.3Privatemode AIReleased Aug 14, 20261,000,000 token context windowIn $1.791 · Out $6.6326
Input
Output
Privatemode AIkimi-k2deprecated
Kimi (latest)
kimi-latestPrivatemode AIReleased Apr 21, 2026256,000 token context windowIn $1.791 · Out $8.9436
Input
Output
Privatemode AIkimi-k2deprecated
Kimi K2.6
kimi-k2.6Privatemode AIReleased Apr 21, 2026256,000 token context windowIn $1.791 · Out $8.9436
Input
Output
Privatemode AIbeta
DeepSeek OCR 2
deepseek-ocr-2Privatemode AIReleased Jan 27, 20268,192 token context windowIn $0.8897 · Out $1.4675
Input
Output
Privatemode AIgpt-oss
gpt-oss-120b
gpt-oss-120bPrivatemode AIReleased Aug 5, 2025128,000 token context windowIn $0.2311 · Out $0.7511
Input
Output
Privatemode AIvoxtral
Voxtral Mini 3B
voxtral-mini-3bPrivatemode AIReleased Jul 1, 202532,000 token context windowSubscription plan pricing
Input
Output
Privatemode AIqwen
Qwen3-Embedding 4B
qwen3-embedding-4bPrivatemode AIReleased Jun 6, 202532,000 token context windowSubscription plan pricing
Input
Output
Privatemode AIwhisper
Whisper large-v3
whisper-large-v3Privatemode AIReleased Oct 1, 2024448 token context windowSubscription plan pricing
Input
Output