Catalog

Threads API

Threads data through one normalized API - per-request USD pricing, no subscriptions, and automatic failover on redundant endpoints.

Endpoints5Providers2From · search$1.20/1k req · GiraffeFull breakdownRead docs
Uptime
100.00%
30d · 483 calls
Requests
483
30d · weekly, last 12 wks
Response
1.1s
median · 30d

Measured performance

What the Threads API actually does

Every call through AnyAPI over the last 30 days, counted here rather than promised.

100.0%
of calls succeeded (483 calls)
Succeeded on first lane 94.8%Recovered by failover 5.2%
Response time estimate (483 calls across endpoints)
Weighted p50 estimate
1.1s
Request-weighted average of endpoint p50 values, not a combined request percentile
Weighted p99 estimate
4.7s
Request-weighted average of endpoint p99 values, not a combined request percentile

FAQ

About the Threads API

The AnyAPI Threads API returns Threads data as normalized JSON across 5 endpoints: post, profile, search, search users, and user posts. You call one endpoint per job with a single key, and AnyAPI routes each request across 2 sources, falling back automatically when one fails. Pricing is in US dollars, from $1.20 per 1,000 requests on the cheapest endpoint, with no subscription and no monthly minimum. Over the last 30 days, 100.0% of Threads calls through AnyAPI succeeded, with a request-weighted median response time of 1.1 seconds across 483 measured calls.

Pricing is in US dollars, from $1.20 per 1,000 requests on the cheapest endpoint, with no subscription and no monthly minimum. You fund one USD wallet, each call draws it down, and a failed request costs $0.

5 endpoints: post, profile, search, search users, and user posts. Every one takes the same authentication and returns the same normalized response shape, so reaching a second Threads endpoint is a change of URL and nothing else.

Over the last 30 days, 100.0% of Threads calls through AnyAPI succeeded, with a request-weighted median response time of 1.1 seconds across 483 measured calls. These are AnyAPI's own measurements of traffic through the gateway, recomputed continuously, not a published service-level target.