showing active exposure to SMH
API use cases
AI market intelligence for financial products
Use CapitalBench data to add AI model positioning, risk appetite, portfolio behavior, live benchmark context, and audit evidence to investor dashboards, trading workflows, research terminals, and quantitative pipelines.
current AI risk appetite, risk-seeking
allocation rows for feature engineering
official result rows with benchmark context
public model profiles with behavior metrics
Financial websites and market data platforms
Add an AI positioning panel to asset pages
Market data sites can show what AI models currently hold beside ETF, sector, theme, or macro pages. The panel works as a context layer next to price, volume, fundamentals, news, and analyst data.
- Models currently holding the asset
- Average active allocation and model count
- Recent allocation changes and related insights
- Live benchmark performance context
/v1/assets/{option_id}/model-holders /v1/positioning/by-asset/{option_id} /v1/insights ETF PAGE SMH - Semiconductors
+2.4% SMH 24.0%
MTUM 10.1%
XLK 6.7%
EWY 6.6%
SPY 6.0%
Brokerage and trading platforms
Place AI context next to the trader's existing workflow
Use CapitalBench as a market-context feed, not an execution instruction. Traders can see consensus, disagreement, risk appetite, and live benchmark movement alongside their own price, volume, macro, and news stack.
- Top consensus assets and categories
- Biggest allocation changes since the prior round
- Live mark-to-market model performance
- Risk appetite split by weekly and monthly tests
/v1/risk-appetite /v1/positioning/active /v1/positioning/changes /v1/live/performance Algorithmic traders and quant teams
Convert AI allocation behavior into research features
Quant teams can ingest raw official-run allocations, return rows, and interim performance history as an alternative data source. CapitalBench does not provide a buy/sell signal; it provides auditable model behavior that can be tested inside your own research process.
AI allocation momentum
/v1/allocations Consensus concentration
/v1/positioning/consensus Risk appetite regime
/v1/risk-appetite Live performance drift
/v1/live/performance/history Benchmark-relative alpha
/v1/returns - CapitalBench API Allocations, returns, live rows
- Feature store Clean, join, version
- Backtest Replay against market data
- Signal ranking Weight by evidence
- Risk controls Turnover, crowding, limits
- Trader UI or strategy input Render or test inside your workflow
CB-2026-06-15-1W
Oracle-relative result
Generated from benchmark math
Weekly and monthly context
/v1/models/behavior /v1/benchmark-evidence /v1/rounds/{round_id}/proof Investment research desks
Track model behavior across market regimes
Research teams can compare frontier model records, concentration, risk style, turnover, peer similarity, and audit evidence without rebuilding CapitalBench's public read model.
- Which models are consistently risk-seeking or defensive
- Which models over-concentrate or crowd into peers
- Which decisions were frozen before prices resolved
- Which benchmark comparisons are equal-run and qualified
Wealthtech and advisor platforms
Explain how AI models are reading the market
Advisor and wealthtech products can use CapitalBench as educational market context: model agreement, current themes, risk appetite, and benchmark evidence. Keep the framing explicit: this is not portfolio advice.
/v1/risk-appetite /v1/positioning/consensus /v1/insights 80 Risk appetite
Semiconductors 24.0%
US Momentum Equities 10.1%
Technology Sector 6.7%
South Korea Equities 6.6%
Educational context from public benchmark records. Not investment advice.
Current live pulse
CB-2026-06-15-1W
Model-level metrics
Round audit links
LLM labs and model providers
Benchmark market behavior, not just market language
Model teams can inspect portfolio decisions, style metrics, risk appetite, turnover, concentration, peer similarity, and resolved performance. The API gives them the public evidence needed to compare model behavior over time.
/v1/models/{model_id}/style /v1/models/{model_id}/behavior /v1/models/{model_id}/portfolios /v1/leaderboards/benchmark-sets Integration pattern
A clean data layer for product teams
CapitalBench works best as an additional intelligence layer: ingest structured API rows, cache them in your data layer, and render them as context in the product surfaces your users already trust.
- 01 Ingest
Pull allocations, returns, risk appetite, model behavior, and proof records.
- 02 Join
Map CapitalBench assets to your ETF, sector, and watchlist identifiers.
- 03 Render
Show AI positioning, consensus, disagreement, and audit links inside your workflow.
- 04 Test
Validate whether model behavior adds value inside your own process.
CapitalBench is benchmark and research infrastructure, not investment advice or a trading recommendation.
Rounds connect frozen prompts, model portfolios, prices, results, and proof metadata.
Public API views filter to official records so retries do not contaminate benchmark data.
Add AI market positioning to your product
Use the CapitalBench API to power investor panels, trading context, quant features, research terminals, and model evaluation dashboards.