Verso vs
SimpleFunctions.

SimpleFunctions' thesis system takes any plain-English sentence and decomposes it into a causal tree of testable sub-claims. Each node receives a probability, and an evaluation heartbeat continuously scans news, refreshes prices, checks milestones, runs an LLM evaluation, and updates confidence. You can inject signals via /api/thesis/{id}/signal and fork public theses. Verso surfaces GPT-5-backed forecast scores on its dashboard but does not expose a structured thesis-decomposition API or a causal-tree model — the two approaches serve fundamentally different workflows.

SimpleFunctions ships Portfolio Autopilot, an autonomous trading agent that uses a 1M-context LLM and 13 data sources. Before any execution, the system passes through a 7-gate risk cascade: kill switch, position limits, drawdown gate, regime check, and additional guardrails. Verso is positioned as a read-only analytics terminal for human traders and does not document an autonomous execution layer. Autopilot is an SF-specific capability with no equivalent in Verso's published product.

Yes. SF ships a 56-tool MCP server covering prices, orderbook depth, thesis creation, signal injection, cross-venue arbitrage scanning, and autonomous trade intents. Add it with one command: `claude mcp add simplefunctions --url https://simplefunctions.dev/api/mcp/mcp`. It works with Claude Code, Cursor, and any MCP-compatible client. Verso does not publish an MCP server or documented programmatic API; all access is through its web terminal UI.

Yes — SF exposes a full REST API at /openapi.json, a CLI with 60+ commands (`npm i -g @spfunctions/cli`), and an MCP server with 56 tools. Public endpoints for prices, orderbook depth, cross-venue pairs, indicators, and the world snapshot require no authentication. Verso does not publish a REST API or SDK; all access is through its web terminal, which makes it unsuitable for programmatic or agent-driven integrations.

SF publishes live Brier scores at /api/calibration, broken down by venue (Kalshi 0.20, Polymarket 0.12 on T-24h price), category, and price bucket over the past 90 days — verifiable with a single curl call. Verso self-reports 73% forecast accuracy on its dashboard. The two metrics are not directly comparable — Brier-based calibration and directional accuracy rates measure different things — but SF's numbers are independently auditable by anyone.

SF pre-computes six indicators across 48K+ active contracts: IY (implied yield), CRI (cliff risk index), LAS (liquidity availability score), EE (event overround), τ-days (time to settlement), and regime label (adverse-selection classification). These are surfaced at /screen and accessible via the API and MCP tools. Verso focuses on price charts and AI forecasts within its dashboard; it does not document pre-computed derivative indicators available for programmatic consumption.

Not for every use case. Verso is a UI-first terminal designed for professional and institutional human traders who want a Bloomberg-style cockpit with GPT-5 forecasts and seat-based access. SF is API and MCP-first, built for developers and agent builders. If your workflow is primarily dashboard-driven without programmatic access needs, Verso may be the better fit. If you are building agents, research pipelines, or automated trading systems, SF's thesis API, autopilot, MCP server, and calibration layer address a different set of requirements.

GET /api/public/cross-venue/pairs?preset=arb returns normalised pairs where the same underlying event is priced differently on Kalshi and Polymarket, along with spread, implied yield, and a conviction score — computed continuously across 48K+ contracts. You can filter by preset (arb, corr) and consume results in agent loops or the CLI. Verso covers both venues in its UI but does not document a dedicated cross-venue arbitrage scanning endpoint or programmatic access to matched pairs.