A credit is the per-call billing unit on Hive. By default, one tool call consumes one credit — whether it's get_price, get_token_security, get_funding_rates, or any other tool. A few high-cost endpoints (large historical exports, complex aggregations) may consume more than one credit per call; the response envelope reports the credits consumed per request.

What is a rate limit?

A rate limit is the maximum requests an API key can send per minute. Hive Demo allows 30 req/min; Analyst is 500 req/min; Pro is 1,000 req/min; Enterprise scales to 3,000 req/min. Hitting the limit returns HTTP 429 with a Retry-After header. Rate limits prevent any single key from monopolizing upstream provider capacity.

Why two limits — monthly credits AND per-minute rate?

Credits cap total monthly volume (cost protection — runaway loops can't bankrupt you). Rate limits cap instantaneous throughput (fairness — one key can't monopolize the queue). An agent doing 100K thoughtful calls/month at 5 calls/min consumes the same credits as one doing 100K calls in a panic-loop at 1,000 calls/min, but the second pattern degrades service for everyone, so rate limits prevent it independently of credit balance.

How are Hive credits different from upstream provider credits?

Hive abstracts away the per-provider credit accounting. One Hive credit equals one call to a Hive tool, regardless of how many upstream provider calls Hive makes to fulfill it. A get_token_security call might internally fan out to GoPlus + DefiLlama + Moralis — each with its own provider credit cost — but the agent sees one Hive credit consumed. Without this abstraction, agents would have to reason about per-provider economics on every call.

What are credits and rate limits? | Hive API Pricing Definition

Credits are Hive's per-call billing unit; rate limits cap requests per minute. Together they define how much an agent can do per month and how fast.

Every Hive tool call consumes credits from your monthly budget and is subject to your plan's per-minute rate limit. The two limits operate independently — a Demo agent can use up its 10,000 credits in 100 quick bursts of 100 calls or in a steady drip over the month, but it cannot exceed 30 calls in any given minute.

Plan ladder (current as of May 2026):

Plan	Monthly credits	Rate limit
Demo	10,000	30 req/min
Analyst	500,000	500 req/min
Pro	2,000,000	1,000 req/min
Enterprise	Unlimited	3,000 req/min

Every plan includes the full tool catalog. Tiers differ only in volume and throughput.

Why two limits

Credits protect cost — a runaway loop can't bankrupt you because the credit cap is per month. Rate limits protect fairness — no single key can saturate the upstream provider queue, even with credits remaining.

Hive abstracts away per-provider credit accounting. One Hive credit equals one Hive tool call, regardless of how many upstream provider calls Hive makes internally to fulfill it. A get_token_security call may fan out to GoPlus + DefiLlama + Moralis behind the scenes; the agent sees one Hive credit consumed. Single-provider APIs like CoinGecko Pro or CoinMarketCap require agents to track per-provider economics on every call.

Why two limits

Related terms