/ UsageChecking status
Status
Service health and organization activity.
Organization activity
All membersRequest activity
Requests / 5 minRequests in 5-minute intervals.
Tokenizer cache
Last hourHit ratio per layer, separate from GPU prefix caching.
Latency
Mean duration- Edge time to first content
- —
- Edge full request duration
- —
- Engine time to first token
- —
- Engine time per output token
- —
Edge: retained requests. Engine: latest scrape interval.
Infrastructure
Kimi K3SMGEC2 gateway
TokenSpeed2 nodes · 8 × GB300 · TP8
Private tunnel between gateway and engine
Awaiting deployment information
GPU utilization and KV occupancy are unavailable.
Base URL OpenAI compatible
API keys
Manage API keys for your account. Keep your keys private and store them securely.
Use an environment variable when connecting your application.
| Name | Secret key | Created | Expires | Created by | Status | Actions |
|---|
Quickstart
Use your API key with the OpenAI Python SDK.
Usage
Your account
Requests
Inference API
5-minute intervals · Your API keys
Requests
Latest 200| Created | Model | Status | Input tokens | Output tokens | Cached tokens | Duration |
|---|
About this data
Up to 7 days or 50,000 service requests are retained. Unavailable values appear as —.
Only request metadata is stored. Input and output content is not saved.
Chat
Settings
Organization access and service configuration.
Deployments
Not configuredNightly releases become available after the first serving release and rollback are verified.
Organization API keys
Organization access| Owner | Name | Secret key | Status | Actions |
|---|
Organization usage
Last 7 days| Time | User ID | Outcome | Input tokens | Output tokens |
|---|