Pull price history into pandas - Ornn Data

This guide loads an H100 SXM price history into a pandas DataFrame, ready for analysis or plotting.

Prerequisites

An API key.
pip install requests pandas

Fetch and load

The /history-simple endpoint returns a clean timestamp / index_value series, ideal for a DataFrame.

import requests
import pandas as pd

API_KEY = "YOUR_API_KEY"
BASE = "https://api.ornnai.com"

resp = requests.get(
    f"{BASE}/api/gpu/H100 SXM/history-simple",
    params={
        "startDate": "2026-05-01",
        "endDate": "2026-05-31",
        "granularity": "daily",  # one point per day
        "limit": 500,            # default is 100, raise it for long ranges
    },
    headers={"Authorization": f"Bearer {API_KEY}"},
)
resp.raise_for_status()

df = pd.DataFrame(resp.json()["data"])
df["timestamp"] = pd.to_datetime(df["timestamp"])
df = df.sort_values("timestamp").set_index("timestamp")

print(df.head())

Output

                           index_value
timestamp
2026-01-01 00:00:00+00:00         2.00
2026-01-02 00:00:00+00:00         2.00
2026-01-03 00:00:00+00:00         2.00
2026-01-04 00:00:00+00:00         2.00
2026-01-05 00:00:00+00:00         2.00

The API returns points most-recent-first. Sorting by timestamp (as above) puts them in chronological order before any rolling calculations or plotting.

Plot it

import matplotlib.pyplot as plt

df["index_value"].plot(title="H100 SXM price index (USD/hr)")
plt.ylabel("USD per hour")
plt.show()

H100 SXM price index line chart for May 2026 — H100 SXM daily price index, May 2026

H100 SXM price index line chart for May 2026 — H100 SXM daily price index, May 2026

Compare multiple GPUs

gpus = ["H100 SXM", "H200", "B200"]
frames = []

for gpu in gpus:
    r = requests.get(
        f"{BASE}/api/gpu/{gpu}/history-simple",
        params={"startDate": "2026-05-01", "endDate": "2026-05-31",
                "granularity": "daily", "limit": 500},
        headers={"Authorization": f"Bearer {API_KEY}"},
    )
    s = pd.DataFrame(r.json()["data"])
    s["timestamp"] = pd.to_datetime(s["timestamp"])
    frames.append(s.set_index("timestamp")["index_value"].rename(gpu))

prices = pd.concat(frames, axis=1).sort_index()
prices.plot(title="GPU price index comparison")

GPU price index comparison line chart — H100 SXM, H200, and B200 daily price index, May 2026

GPU price index comparison line chart — H100 SXM, H200, and B200 daily price index, May 2026