Overview

TypeScript SDK for Thetanuts Finance V4 — options trading on EVM chains.

Using an LLM (Claude, Cursor, ChatGPT)? This SDK ships an MCP server that exposes ~100 tools — read state, build transactions, run pricing math. One-line install: npx -y @thetanuts-finance/mcp, or paste the LLM context prompt into your LLM. Want to actually trade from chat? Pair it with Base MCP and approve each transaction in your wallet (Trade from Chat), or run a fully autonomous agent with its own wallet (AgentKit). Start at AI Agents — Pick Your Route.

Features

Complete Options Trading: Fill orders, manage positions, handle RFQs
Settlement Types: Cash-settled and physically settled options
Multi-Strategy Support: Spreads, butterflies, condors, iron condors (cash-settled)
Real-time Data: WebSocket subscriptions for live updates
Type-Safe: Full TypeScript support with comprehensive type definitions
Modular Design: Use only what you need
ESM and CJS builds: Maximum compatibility

OptionBook vs RFQ

The SDK supports two trading systems. Choose based on your use case:

OptionBook

RFQ (Factory)

What

Fill existing market-maker orders

Create custom options via sealed-bid auction

When to use

Quick trades on listed options (vanilla and multi-leg)

Custom options via sealed-bid auction: any strike, any expiry, cash-settled or physically settled

Structures

Vanilla, spread, butterfly, condor, iron condor (cash-settled)

Vanilla, spread, butterfly, condor, iron condor (cash-settled or physically settled for vanilla)

Key methods

fillOrder(), previewFillOrder()

buildRFQRequest(), requestForQuotation()

Pricing

Order prices from fetchOrders()

MM pricing from getAllPricing()

Data source

Book indexer (/api/v1/book/)

Factory indexer (/api/v1/factory/)

User data

getUserPositionsFromIndexer()

getUserRfqs(), getUserOptionsFromRfq()

Stats

getBookProtocolStats(), getBookDailyStats()

getFactoryProtocolStats(), getFactoryDailyStats()

Collateral

Paid upfront by taker

collateralAmount = 0 (held by factory)

Settlement

Cash-settled (payout in USDC/WETH/cbBTC based on price difference at expiry)

Cash-settled by default; physically settled optional via buildPhysicalOptionRFQ()

Which one should I use?

Both OptionBook and RFQ create cash-settled options using the same on-chain contracts. The difference is how you get there:

Is there an existing order that matches your trade?
  YES --> Use OptionBook (instant fill, no waiting)
  NO  --> Use RFQ (custom strike/expiry, MMs compete on price)

Need physical settlement (actual token delivery at expiry)?
  YES --> Use RFQ with buildPhysicalOptionRFQ()
  NO  --> Either works. OptionBook is faster if an order exists.

OptionBook is like a limit order book. Makers have already posted orders with set strikes, expiries, and prices. You pick one and fill it instantly.

RFQ is like sending out a request for bids. You specify what you want (any strike, any expiry), and market makers compete via sealed-bid auction. It takes ~60 seconds for offers to arrive, but you get competitive pricing on exactly the parameters you need.

Features

OptionBook vs RFQ

Which one should I use?

See also