spareapi /// 500+ APIs · cheapest in market · 30s setup

one key
every API
for AI agents

One key, pay-per-call, every model and every web API your agent needs — social, search, crypto, scrape, OTP, browser, even other agents. Plug into Claude, Cursor, Codex, Gemini in 30 seconds.

500+
APIs · 1 key
~30ms
p50 gateway overhead
$0.92
avg agent / month
spareapi ~/agent · v2.3
live
// 01 Teach your agent the spare skill
_
01 the marketplace

Production APIs
agent-ready

Twitter API
SOCIAL
Full Twitter/X graph — tweet details, user lookup, search, followers, media. Multi-region failover, sub-second p50.
1.1M calls·576ms·99.17%
Serper API
INFRASTRUCTURE
Google Search results as JSON. Organic, ads, places, news, shopping, scholar. Refreshed every 30 minutes.
87.2k calls·760ms·98.82%
Claude Router
AI MODELS
Anthropic native format (/messages) + OpenAI ChatCompletions adapter on the same endpoint. Falls over to Gemini if Anthropic 5xx.
8.1k calls·2.4s·100%
OpenRouter
AI MODELS
Unified gateway for 200+ open and closed LLMs. Pick by price, latency, or capability — we route automatically.
556 calls·17s·100%
Reddit
SOCIAL
Subreddit feeds, comments, search, modlogs. Honors the official rate-limit budget for you. Best-in-class for community-trend agents.
2.4k calls·490ms·99.6%
TikTok Proxy
SOCIAL
Trending feed, video metadata, hashtag stats, user profiles. Built for content-analysis agents.
330 calls·3.9s·83.94%
02 the proposition

One key
every model, every API,
pay per call

Top up once. Spend a fraction of a cent per call on any provider we route to. No min commit, no per-provider account, no overage trauma.

Claude
Cursor
Codex
Gemini
+any agent
SpareAPI
Twitter
TikTok
Reddit
OpenAI
Google
Chains
+ 500 more APIs
500+ APIsp50 ~30msno min commit
03 capabilities

Plug-and-play
no glue code

Cross-API skills — each capability hides routing, auth, retries, and schema-translation behind a single endpoint. Drop one ID in your tool list, spare handles the rest.

AI / Search / Social / On-chain
call / render / settle
auth / logs / quotas / discovery
spare/services
28 actions live
$ npx spareai list --format table
CATEGORYACTIONS
Social9
twitter.tweet_detailtwitter.user_by_screen_nametwitter.user_tweets+6 more
Web9
web.searchweb.search_realtimeweb.search_video+2 more
AI-Services7
ai.text.chat.fastai.text.chat.reasoningai.text.chat.auto+1 more
Crypto2
crypto.token.pricecrypto.token.metadata
Platform1
task.poll
04 get started

Two steps
and your agent calls a real API

01 /// install the skill

One skill that hands your AI the live spare tool catalog. Works with Claude Code, Cursor, Codex CLI, Gemini CLI, and any MCP-compatible harness.

$npx skills add spareapi
skill is open source on github: spareapi ↗
02 /// use it

Type /spareai followed by what you want. The skill picks the right action, fills params, and routes through spare — usage is metered to your key.

/spareai find tweets from @elonmusk this week
/spareai google scholar "agent eval benchmarks 2026"
/spareai rent a US number to verify an OpenAI signup
/spareai ask a code-review agent to look at this PR
05 pricing

Pay per call
cheaper than holding 12 accounts

Callsparedirect / othersdelta
Twitter detail$0.0024$0.01280%
Web search$0.0010$0.005080%
Claude Sonnet · 1k tok$0.0030$0.0030at cost
GPT-4o mini · 1k tok$0.0001$0.0001533%
SMS OTP receive$0.0080$0.02060%
Render & extract page$0.0012$0.00885%
measured 2026-05 · median of 1k calls per row · full methodology at /pricing/methodology
ready when you are

Get an API key
or publish your first endpoint