Pricing and limits

Every feature is on every plan, Free included. Plans differ only in how much you use and how soon you receive the live record. Flat fees, never per trade.

On this page

Every plan includes: every market and its history, the full settled record, the consolidated prices, Langbot Sim on any recorded event, the live record over WebSocket, SSE, REST and signed webhooks, receipts, the MCP server and every tool, the SDKs and the CLI. Plans differ only in usage and in the delay on the live record. Fees are flat: never per trade, per contract or per volume.

What is on sale today, and the current prices, are on langbot.ai/pricing.

	Free	Builder	Standard
Price	$0	$49 / month	$199 / month
Delay on the live record	+60 s	+10 s	+2 s
API requests	60 / min	600 / min	3,000 / min
Concurrent streams (WebSocket + SSE)	1	5	20
Subscriptions	5	25	200
API tokens	5	20	100
Webhook endpoints	1	5	unlimited
Webhook deliveries	100 / day	10,000 / day	unlimited
History queries (full record)	500 / day	unlimited	unlimited
Model calls	100 / day	unlimited	unlimited
Sim runs (any recorded event)	20 / month	100 / month	unlimited
Consolidated prices	20 / min	120 / min	600 / min

Anonymous calls to public endpoints are served like Free, with the Free request budget shared per client IP. Real time, private sources and redistribution are custom contracts: write to hello@langbot.ai.

How allowances work

Days and months are UTC; allowances reset at 00:00 UTC (the first of the month for Sim runs).
A history query is one call to /v1/history/*, /v1/context/*, /v1/markets/{ticker}/brief or /v1/events/{id}/replay. A model call is one call to /v1/models/*. A Sim run is one POST /v1/sim/runs, or one local langbot sim run (it reads the event's timeline once).
A spent daily or monthly allowance answers 429 with code: quota_exceeded and quota: {dimension, plan, used, limit, period, resets_at}. A held count at its cap (subscriptions, tokens, webhook endpoints) answers 402 with the same code.
GET /v1/me returns what you have used of each allowance, and /.well-known/langbot-pricing.json (or the MCP tool get_pricing) publishes the plans as data.

Details and every error code: REST API.

The delay

The delay is a policy applied per plan, not a different pipeline: every plan receives the same record, held back by its delay. t_emitted is when Langbot sent the event; the stream's hello frame and every REST response tell you the delay you are on (delay_ms, X-Langbot-Delay-Ms).

Billing

Monthly, cancel any time from App → Billing. Prices in USD; taxes where applicable. A lapsed subscription is served as Free until it is renewed.