01
Format Translation
Use the same key across /v1/messages or /v1/chat/completions. The gateway seamlessly transforms requests to match your upstream provider.
v1 · Checking status...
ONE KEY · THREE FORMATS · TOTAL CONTROL
The most capable thinking model,
your API gateway.
Power Claude Code, Cursor, and your custom apps with Claude Fable 5 Thinking. Choose Anthropic, OpenAI, or NVIDIA compatible formats; manage usage from a single dashboard.
99.95% target uptime
0 prompt logs stored
3 compatible formats
claude-code · CoreQM proxy
# Configure environment
export ANTHROPIC_BASE_URL="https://api.yourdomain.com"
export ANTHROPIC_AUTH_TOKEN="sk-live-••••••••••••"
export ANTHROPIC_MODEL="claude-fable-5-thinking"
# Launch Claude Code directly
$ claude
✓ Connected to CoreQM Gateway
Model: claude-fable-5-thinking (active)
Streaming: ready · latency: 82ms/ MODELS
One model.
Works everywhere.
The public model name stays constant; backend providers, pricing, and formats can be tuned at any time.
/ WHY COREQM
02
Transparent Usage
Daily and total tokens, credit balance, request history, and expiration in one view. Configurable markup for resale.
03
Key Security
Sub-keys are displayed only once. Only cryptographic SHA-256 hashes are stored in the database with instant revocation.
04
Multi-Provider Routing
Store encrypted upstream API keys for Claude, OpenAI, NVIDIA, and Gemini with granular user-group routing.
/ PACKAGES
Day & token based.
No surprises.
Instant top-up after order review. Once approved, credits are added directly to your account balance.
/ KEY STATUS
How is your key performing?
Check expiration date, daily usage, and limits securely without logging in. Key values are never logged or stored.
/ DEVELOPER DOCS
Change the Base URL.
Your code works as-is.
Connect Claude Fable 5 Thinking seamlessly to your favorite tools and libraries.
/ FREQUENTLY ASKED
Clear answers before
you get started.
Quick answers regarding accounts, security, and protocol compatibility.
Do I need to modify my existing Claude Code setup?
No. You only need to configure the API key, base URL, and model environment variables.
Are my prompts and completions stored?
By default, prompt content is never logged. Only billing metadata like token counts, model names, latency, and status are kept.
Can I recover my API key if lost?
For security reasons, no. You can revoke the old key and generate a fresh one in seconds.
Can I use the same key for both OpenAI and Anthropic endpoints?
Yes. The key remains identical; simply change the target endpoint. The gateway handles protocol and body conversion.
Ready to get started
Generate your first API key