Decision ModelsGold
XAU / USD
—
—Collecting
Waiting for a quote
Metals Pulse
Gold and silver prices.
Connecting
Waiting for prices.
Loading prices.
USD / troy oz
Silver
XAG / USD
—
—Collecting
Waiting for a quote
Reference prices. Dealer prices may differ.
AI assessments
Price changes, not predictions.
Awaiting assessment
Waiting for history
Need more history
An assessment will appear when enough prices have been collected.
- Period
- Not available
- Change
- —
ActionNeed more data
No assessment yet
Inspect assessment
No assessment available.
Evidence
- Waiting for price history.
Returns by period
These probabilities describe the model's choices. They are not calibrated confidence or forecasts.
Awaiting assessment
Waiting for history
Need more history
An assessment will appear when enough prices have been collected.
- Period
- Not available
- Change
- —
ActionNeed more data
No assessment yet
Inspect assessment
No assessment available.
Evidence
- Waiting for price history.
Returns by period
These probabilities describe the model's choices. They are not calibrated confidence or forecasts.
Price history
No price history yet
The chart will fill as prices are collected.
No observed history yet0 points
Only stored prices are shown. Chart controls do not change the AI assessment.
Gold change—
Silver change—
Period shown—
Sample interval—
Data and costs
Stored history
Prices are retained to build longer comparisons.
- Raw prices
- — retained
- 5-minute samples
- — retained
- Hourly samples
- — retained
History startsWaiting for the first observation
These are retention limits, not available coverage. Older prices are not backfilled. Longer ranges use the last price collected in each interval. Sampled highs and lows are not exchange highs and lows.
How it works
Each metal is assessed from its own price history. The assessment uses the longest reliable period available, up to one day. A five-minute assessment can appear first; longer periods become available as prices are collected.
AI chooses among patterns supported by the measurements. “Review” means take a closer look, not buy or sell. The assessment does not predict prices or infer the cause of a move.
Assessment rules
The assessment considers observed returns, direction consistency, sampled ranges, pullbacks, rebounds and variation. It uses the longest reliable window up to one day, with weekly history as context when available.
Experimental attention floors are 0.03% over five minutes, 0.10% over an hour, 0.25% over six hours, 0.50% over a day and 1% over a week. Sustained direction also requires consistent movement and limited retracement; pullback and rebound assessments require an earlier move within the observed period.
When AI is unavailable, a rules-based assessment may be shown and is labelled “Rules”. Probabilities describe model choices, not a validated forecast.
Price source
- Price source
- Loading source
- Collection interval
- —
- Continuous recent history
- —
- Last collection
- —
Source timestamps record when Gold API fetched a price, not an exchange trade. The upstream venues are unspecified; these are USD per troy ounce spot references, not executable buy or sell quotes.
Stale or missing data pauses the current assessment. Gaps are marked on the chart. Unchanged prices do not establish whether a market is open.
Source methodology ↗