OpenAI-compatible · Private by design

Cipher API

The same private model behind Secrypt chat — Cipher — as an OpenAI-shaped API. Hook it into Cursor, Continue, VS Code, scripts, and agents.

Pricing (API ≠ chat)

Chat subscriptions cover the Secrypt UI. Cipher API is usage-based: a daily allowance, then prepaid credits.

Chat overage (UI) is $0.05/message. API overage is always $0.01/request. Buy credits from Account. See /pricing.

Base URL

https://secrypt.space/v1

Model IDs: cipher (preferred) or secrypt-cloud (alias).

Cursor cookbook

In Cursor Settings → Models → OpenAI API Key / Override:

• OpenAI API Key — paste your sk_sec_… key
• Override OpenAI Base URL — https://secrypt.space/v1
• Model — cipher

# ~/.cursor/ or Cursor Settings → Models
OpenAI Base URL: https://secrypt.space/v1
OpenAI API Key:  sk_sec_YOUR_KEY
Model:           cipher

Continue (VS Code)

{
  "models": [{
    "title": "Cipher",
    "provider": "openai",
    "model": "cipher",
    "apiBase": "https://secrypt.space/v1",
    "apiKey": "sk_sec_YOUR_KEY"
  }]
}

cURL

curl https://secrypt.space/v1/chat/completions \
  -H "Authorization: Bearer sk_sec_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cipher",
    "messages": [{"role":"user","content":"Say hello in one sentence."}]
  }'

Python (OpenAI SDK)

from openai import OpenAI
client = OpenAI(
    api_key="sk_sec_YOUR_KEY",
    base_url="https://secrypt.space/v1",
)
r = client.chat.completions.create(
    model="cipher",
    messages=[{"role": "user", "content": "Ping."}],
)
print(r.choices[0].message.content)

Limits & privacy

• Keys are hashed at rest; revoke anytime from Account.
• API usage is metered separately from chat.
• Optional web tools follow your account/search settings on the server.
• Two concurrent slots on the current GPU worker — under heavy load, requests may queue briefly.