| Article | Analyzed | Src | .md | Nodes | Linked | New N | Edges | New E | Tokens | Est. USD |
|---|
| Date & time | Prompt (chars) | Request analysis | Tokens in | Tokens out | Est. USD | React. |
|---|
| ID | Element | Status | Original # | Match | Tokens (in/out) | Est. USD | Result |
|---|
A/H/D overlay
Accuracy
ID match
Result
| Time | Call | Model | Prompt | Candidates | Total | Est. (USD) |
|---|
| Process | Nodes checked | N before | N after | E before | E after | Tokens | Est. USD | Note |
|---|
API keys
Create named keys for external integrations.
Usage rule: all API calls must be logged in Token usage with the call prefix
api · .
Key details
Per-key usage counters and total spend.
API usage guide (English)
Base URL: http://195.201.173.59/
Authentication
Send API key in header:
Authorization: Bearer <YOUR_API_KEY>
1) Text answer + tokens + cost
POST /api/v1/query/text
{
"question": "Your question",
"mode": "chat"
}
Response fields only: ok, text (final answer), promptTokens,
candidatesTokens, totalTokens, costUsd. No request ids, model
names, or step-by-step usage breakdown.
2) Voice answer + tokens + cost
POST /api/v1/query/voice
{
"question": "Your question",
"voice": "default",
"format": "mp3"
}
Same as text, plus audioBase64 and audioMimeType (MP3). Totals include
speech synthesis.
3) Token statistics by period
GET /api/v1/stats/tokens?range=all|30d|24h|60m
Returns request count, input/output/total tokens, and total estimated USD for selected period.
4) Memory Tree structure + request counters
GET /api/v1/stats/memory-tree
Returns nodes, edges, articles, and total chat requests (User's chat + AI chat checker).
Logging requirement
Every API request must be written to Token usage with call prefix
api · followed by endpoint/action.