Matchr vs
SimpleFunctions.

Matchr's market search finds the best-price entry for a given event across 1,500+ markets. SimpleFunctions' thesis system starts from a natural-language sentence — a claim like 'Taiwan chip exports will fall in H2 2026' — and decomposes it into a causal tree of testable sub-claims. Each sub-claim is matched to live contracts on Kalshi and Polymarket, probabilities are propagated across the tree, and an autonomous evaluation heartbeat runs on a schedule: news scan, price refresh, milestone check, LLM confidence update. Signals can also be injected manually via /api/thesis/{id}/signal. No current prediction market data product offers this.

SimpleFunctions indexes 48K+ active contracts across Kalshi and Polymarket. Matchr states coverage of 1,500+ markets across multiple venues. The difference partly reflects how contracts are counted — SF counts every sub-contract (e.g. each strike in a resolution ladder), while aggregators often count by parent event. SF covers the same primary venues Matchr uses. If you specifically need markets beyond Kalshi and Polymarket, verify Matchr's full venue list against your target market before choosing.

Matchr mentions automated yield strategies as a product feature but does not publish details of its architecture or risk management in public materials. SF's Portfolio Autopilot is an autonomous agent: a 1M-context LLM with access to 13 data sources evaluates each trade idea, then passes it through a 7-gate risk cascade — kill switch, position limits, drawdown gate, regime check, and others — before any execution. Every decision and gate outcome is logged. The Autopilot is configurable, auditable, and can be halted via a single kill switch.

SF pre-computes six indicators across 48K+ contracts: IY (implied yield), CRI (cliff risk index, measuring price sensitivity to near-term resolution events), LAS (liquidity availability score), EE (event overround, measuring market maker margin), τ-days (time to settlement in calendar days), and a regime label classifying each contract's adverse-selection environment. These are available at /screen and callable via MCP tool invocations. None require manual computation; they update continuously as prices and liquidity change.

Yes. SF ships a 56-tool MCP server accessible by running claude mcp add simplefunctions --url https://simplefunctions.dev/api/mcp/mcp. Once added, any MCP-compatible client — Claude Code, Cursor, or others — can call SF tools directly: search markets, fetch orderbook depth, create theses, read calibration scores, pull trade ideas, or query the world snapshot. No additional authentication is needed for read-only tool calls. Matchr does not publish a comparable MCP integration.

Run curl https://simplefunctions.dev/api/calibration. The endpoint returns Brier scores broken down by venue (Kalshi, Polymarket), category, and price bucket, computed over the past 90 days against T-24h prices. Current values: Kalshi 0.20, Polymarket 0.12. Brier scores range from 0 (perfect) to 1 (worst); lower is better. The data is live and re-computed on each request, so the number on this page and the number from your terminal will match.

Based on Matchr's public website, its product is presented as a web interface for market search and smart routing. No public REST API documentation or developer portal appears in Matchr's public materials. If you need programmatic access to cross-venue prediction market data, SF exposes a fully documented REST API at simplefunctions.dev/openapi.json, a 56-tool MCP server, and a CLI installable via npm i -g @spfunctions/cli — all usable without authentication for read operations.