TokenBridge API Reference

01Authentication

Every request to the proxy gateway and account endpoints requires a TokenBridge API key passed as a Bearer token.

bashAuthorization: Bearer tb_live_xxxxxxxxxxxxxxxxxxxx

API keys are prefixed tb_live_. Obtain yours at /signup or from your dashboard.

Rate-limit response headers

Every proxy response includes these headers:

Header	Description
X-RateLimit-Limit	Maximum requests per 60-second window (currently 100)
X-RateLimit-Remaining	Requests remaining in the current window
X-RateLimit-Reset	Unix timestamp (seconds) when the current window resets

02POST/api/proxy/v1/chat/completions

OpenAI-compatible chat completions endpoint. Supports streaming via SSE. Deducts from your EUR wallet on every successful call.

Request

bashcurl https://tokenbridge-zoos.polsia.app/api/proxy/v1/chat/completions \
  -H "Authorization: Bearer tb_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-chat",
    "messages": [{"role": "user", "content": "Hello, world!"}]
  }'

Request body

Field	Type	Required	Description
model	string	Yes	Model ID (see table below). OpenAI aliases gpt-4o, gpt-4o-mini, gpt-3.5-turbo map to deepseek-chat.
messages	array	Yes	Array of { role, content } objects. At least one message required.
stream	boolean	No	Set to true to receive a Server-Sent Events stream. Billing still applies.
temperature, top_p, max_tokens, ...	any	No	Standard OpenAI chat parameters are forwarded to the upstream provider as-is.

Supported models

Model	Input (€/1M tokens)	Output (€/1M tokens)	Provider
deepseek-chat	€0.18	€0.36	DeepSeek
deepseek-reasoner	€0.44	€0.88	DeepSeek
doubao-pro-32k	€0.14	€0.37	ByteDance Volcano
doubao-lite-32k	€0.10	€0.28	ByteDance Volcano
qwen-max	€2.06	€8.26	Alibaba DashScope
qwen-plus	€0.51	€1.55	Alibaba DashScope
qwen-turbo	€0.13	€0.39	Alibaba DashScope

Sample response (non-streaming)

json{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1750000000,
  "model": "deepseek-chat",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! How can I help you today?"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 10,
    "completion_tokens": 9,
    "total_tokens": 19
  }
}

Rate limit

100 requests/minute per API key. Exceeding the limit returns HTTP 429 with a retryAfterMs field indicating the milliseconds until the window resets.

Error codes

03GET/api/proxy/v1/models

Lists all supported models in OpenAI-compatible format. Auth required (no billing deduction).

bashcurl https://tokenbridge-zoos.polsia.app/api/proxy/v1/models \
  -H "Authorization: Bearer tb_live_YOUR_KEY"

Sample response

json{
  "object": "list",
  "data": [
    {
      "id": "deepseek-chat",
      "object": "model",
      "created": 1748260000,
      "owned_by": "deepseek"
    },
    {
      "id": "deepseek-reasoner",
      "object": "model",
      "created": 1748260000,
      "owned_by": "deepseek"
    }
  ]
}

The data array contains all 7 supported models. Pass the id field verbatim as the model in chat completions requests.

04GET/api/me/usage

Returns current wallet balance and the last 50 usage events for the authenticated API key. Auth: Bearer tb_live_... header only (no cookie).

bashcurl https://tokenbridge-zoos.polsia.app/api/me/usage \
  -H "Authorization: Bearer tb_live_YOUR_KEY"

Sample response

json{
  "balance_eur_cents": 4800,
  "balance_eur": "48.00",
  "usage": [
    {
      "model": "deepseek-chat",
      "input_tokens": 250,
      "output_tokens": 180,
      "cost_usd": 0,
      "created_at": "2026-07-25T14:32:10.000Z"
    }
  ]
}

The usage array contains rows from the legacy usage_log table. For the canonical EUR-denominated ledger with reasoning_tokens and eur_cost, use GET /api/usage/history.

05GET/api/usage/history

Itemised per-call rows from the canonical usage_events ledger with full EUR cost data. Auth: tb_session cookie or Bearer header.

Query parameters

Param	Format	Default	Description
from	YYYY-MM-DD	—	Inclusive start date (UTC). Omit for unbounded.
to	YYYY-MM-DD	—	Inclusive end date (UTC). Omit for unbounded.
limit	integer	500	Max rows per page (hard cap: 1000).
offset	integer	0	Row offset for pagination.

bashcurl "https://tokenbridge-zoos.polsia.app/api/usage/history?from=2026-07-01&to=2026-07-25&limit=100" \
  -H "Authorization: Bearer tb_live_YOUR_KEY"

Sample response

json{
  "rows": [
    {
      "created_at": "2026-07-25T14:32:10.000Z",
      "model": "deepseek-chat",
      "input_tokens": 250,
      "output_tokens": 180,
      "reasoning_tokens": 0,
      "eur_cost": 0.0001098,
      "eur_cents": 0,
      "request_id": "req_abc123",
      "status": "completed"
    }
  ],
  "total": {
    "count": 1,
    "total_eur": 0.0001098,
    "eur_cents": 0,
    "eur": "0.00"
  },
  "from": "2026-07-01",
  "to": "2026-07-25",
  "limit": 100,
  "offset": 0
}

06GET/api/usage/export.csv

Downloads usage as a CSV attachment. Same auth as /history. Capped at 10,000 rows per export.

Query parameters

Param	Format	Description
from	YYYY-MM-DD	Inclusive start date
to	YYYY-MM-DD	Inclusive end date

bashcurl "https://tokenbridge-zoos.polsia.app/api/usage/export.csv?from=2026-07-01&to=2026-07-25" \
  -H "Authorization: Bearer tb_live_YOUR_KEY" \
  -o usage.csv

CSV columns

Column	Description
timestamp	ISO-8601 UTC call timestamp
endpoint	Model ID used for the call
prompt_tokens	Input tokens billed
completion_tokens	Output tokens billed
reasoning_tokens	Reasoning tokens (deepseek-reasoner only, billed at output rate)
credits_eur	EUR cost deducted (6 decimal places)
request_id	Opaque request identifier for support queries
status	Always completed for billed rows

07Dashboard usage endpoints

These endpoints power the dashboard UI. Auth: tb_session cookie or Bearer header. Not typically called by production integrations directly.

Path	Params	Response summary
GET /api/usage/recent	limit (max 200, default 50)	{ usage: [...] } — last N calls from usage_log
GET /api/usage/daily	days (max 90, default 30)	{ daily: [...] } — EUR cents per day per model for the stacked bar chart
GET /api/usage/current	none	{ used, limit, pct } — current-month spend vs wallet balance (wallet card progress bar)
GET /api/usage/topups	limit (max 50, default 20)	{ topups: [...] } — last N Stripe wallet top-ups

08POST/api/keys/rotate

Rotates your API key. The old key is invalidated immediately; requests using it will return 401 after rotation. Auth: tb_session cookie or Bearer header.

bashcurl -X POST https://tokenbridge-zoos.polsia.app/api/keys/rotate \
  -H "Authorization: Bearer tb_live_YOUR_KEY"

Response

json{
  "apiKey": "tb_live_YOURNEWKEYHERE"
}

09EUR pricing reference

All costs are deducted from your EUR wallet per call, atomically, rounded up to the nearest EUR cent. No subscriptions, no expiry on credits.

Model	Input (€/1M tokens)	Output (€/1M tokens)	Notes
deepseek-chat	€0.18	€0.36
deepseek-reasoner	€0.44	€0.88	Reasoning tokens billed at output rate
doubao-pro-32k	€0.14	€0.37
doubao-lite-32k	€0.10	€0.28
qwen-max	€2.06	€8.26
qwen-plus	€0.51	€1.55
qwen-turbo	€0.13	€0.39

For deepseek-reasoner, reasoning_tokens are billed at the same rate as output tokens. Cost formula: (input_tokens / 1M × input_rate) + ((output_tokens + reasoning_tokens) / 1M × output_rate), rounded up to nearest cent.

A low-balance email alert fires when your wallet drops below €5. Top up at /buy or from your wallet page.

10/invoices — billing history

The /invoices page lists your Stripe-backed billing history. Auth: session (cookie or Bearer). No API endpoint — rendered as an HTML page from the dashboard.

Path	Auth	Description
GET /invoices	Session	Lists all invoices, newest first (up to 100 rows)
GET /invoices/:id/download.pdf	Session	Downloads a branded PDF for the given invoice row. The :id is the internal database row ID shown in the invoice list.