Use case · Market making

Quote both sides. Collect the spread.

A market maker posts a buy order just below the mid-price and a sell order just above. When a taker hits either side, you collect the spread. Polymarket's reward markets pay LP fees on top. Technically: we maintain inventory targets per market, re-quote on book moves, and stop quoting if inventory drifts outside bands.

50 /mo
"polymarket market maker" US search volume
src: DataForSEO May 2026
0.10-0.50¢
Typical bid/ask spread on liquid Polymarket books
src: 30-day book snapshots
12
Reward markets we actively quote across
src: 5 client deployments
~30s
Median re-quote interval under normal volatility
src: our default config
How it works

Four components, all configurable.

Quoting is the easy part. Knowing when to stop is the hard part.

01 · QUOTE

Post limit orders at mid ± spread, sized against your inventory target. As fills move you long or short, your quotes shift to attract the offsetting flow.

02 · RE-QUOTE

When the book moves more than your tolerance (default 1%), cancel and re-post. Polymarket's CLOB charges nothing for cancels - but timing matters; you'll get adversely selected if you're slow.

03 · INVENTORY-CAP

Each market has a max long and max short. Hit the cap, stop quoting that side, let inventory bleed off naturally.

04 · KILL-SWITCH

If realized P&L drops below your daily floor, every position gets unwound and the bot sleeps. No 'one more trade'.

Config

Two-sided quoting config.

Inventory targets and kill-switch are non-negotiable. We refuse to ship without them.

mm.yaml config
# Market making config - example reward markets
strategy: "two-sided-quote"

markets:
  - id: "btc-15m-mar-2026"
    target_spread_cents: 0.4
    inventory_target_usdc: 0      # market-neutral
    max_long_usdc: 2_500
    max_short_usdc: 2_500

re_quote:
  trigger_book_move_bps: 100     # 1% book move
  min_interval_seconds: 15
  max_quotes_per_minute: 8

risk:
  daily_pnl_floor_usdc: -300    # kill-switch
  max_inventory_pct_of_book: 25  # don't be the whole book
Honest framing

Things to know before you wire funds.

i
Reward markets are where it makes sense.

On a vanilla market, you're earning the spread minus your inventory cost. On reward markets, Polymarket pays you LP fees on top. The math only works on the latter for most retail-sized deployments.

i
Adverse selection is the silent killer.

Informed flow will hit your quote one tick before the market moves against you. Re-quote latency matters. We log per-fill adverse selection and surface it on your dashboard so you can tune.

!
Inventory drift compounds.

A small directional bias in fills over days accumulates into a large net position. Inventory caps + scheduled rebalances are the only defense. We bake both in.

×
Don't quote markets you don't understand.

If you can't explain why the price is where it is, you'll get picked off. We refuse to ship MM bots on markets we can't write a one-paragraph thesis for.

Starting points

Where market making typically works on Polymarket.

Reward markets first. Volatile event markets second. Avoid thin books.

BTC reward markets LP fees + spread

5/15/60-minute BTC threshold markets pay LP fees and have constant flow

5/15/60-minute BTC threshold markets pay LP fees and have constant flow. Our most-deployed strategy.

Major political markets Spread-only

High-attention markets get tight spreads quickly

High-attention markets get tight spreads quickly. MM works pre-event, gets hard near resolution.

Crypto event markets Spread + funding

ETF approvals, halvings, protocol upgrades

ETF approvals, halvings, protocol upgrades. Slow-burn liquidity, occasional volatility spikes - good for patient MM.

Budget bracket

Where this typically lands.

Single-market MM
$15k-$30k · 4-5 weeks

Quote one reward market with inventory caps and kill-switch.

  • One market
  • Inventory-aware quoting
  • Grafana dashboard
  • Discord alerts on kill-switch
  • 60-day warranty
Get started

Ready to provide liquidity?

Tell us your capital, target spread, and which markets you'd like to quote. We'll come back with a backtest and an implementation plan.