Market Data API

Every model on Qubax has a live price that moves with supply. The market API gives you that price, its 24-hour change, high and low, token volume, and full candlestick (OHLC) history. It is the same data behind the Market pages. It is free and needs no API key.

How prices are measured

• Text models: price is a blended USD price per 1M tokens: 75% input + 25% output. This matches the AI Price Index.
• Media models (image, video, audio): unit is set (for example job or second) and the price is per unit.
• benchmark is the OpenRouter price on the same basis, so you can compute the saving yourself.
• A value of null means "not enough data", never zero.

Ticker: all models

GET /v1/market/models. Sorted by 24h token volume. Cached for 60 seconds.

curl https://api.qubax.ai/v1/market/models

{
  "object": "list",
  "data": [
    {
      "id": "deepseek-v4.1-flash",
      "name": "DeepSeek V4.1 Flash",
      "modality": "MULTIMODAL",
      "unit": null,                 // null = price per 1M tokens
      "price": 0.0093,              // blended USD per 1M tokens
      "input_per_1m": 0.0053,
      "output_per_1m": 0.0213,
      "benchmark": 0.057,           // OpenRouter, same basis
      "discount_pct": 83.64,
      "change_24h_pct": 13.81,
      "change_7d_pct": -85.2,
      "high_24h": 0.02625,
      "low_24h": 0.00716,
      "tokens_24h": 2863081647,
      "requests_24h": 15573,
      "spark_7d": [0.06, 0.05, 0.04]  // 28 points, one every 6 hours
    }
  ],
  "updated_at": "2026-10-09T23:13:44.837Z"
}

One model

GET /v1/market/models/{id} adds 30-day change and range, the lowest price ever recorded, 7-day volume, the number of price updates in the last 24 hours, and when history starts.

Candlesticks

GET /v1/market/models/{id}/candles

curl "https://api.qubax.ai/v1/market/models/glm-5.3/candles?interval=1h&series=blended"

{
  "id": "glm-5.3",
  "interval": "1h",
  "series": "blended",
  "unit": null,
  "candles": [
    { "t": 1791579600, "o": 0.3360, "h": 0.4239, "l": 0.3301, "c": 0.3369 }
  ],
  "volume":    [ { "t": 1791579600, "tokens": 15200000, "requests": 120 } ],
  "benchmark": [ { "t": 1791579600, "v": 0.5792 } ]
}

t is the bucket start in Unix seconds (UTC). Days start at 00:00 UTC and weeks start on Monday. Open is the price in force at the start of the bucket, close is the price at its end, and high and low are the extremes reached inside it. A bucket with no price change is a flat candle. Buckets before the first recorded price are left out, never invented.

import requests

r = requests.get("https://api.qubax.ai/v1/market/models/glm-5.3/candles",
                 params={"interval": "1d"}).json()

for c in r["candles"][-7:]:
    print(c["t"], "open", c["o"], "close", c["c"])

Limits and errors

• Responses are cached for 30 to 600 seconds, so polling more often than once a minute gives you the same data.
• 400: invalid model id, interval or series.
• 404: the model is unknown or has no price yet.