A credential per vendor
Each integration brings its own key, its own rate limit, and its own rotation schedule. Onboarding a new source means onboarding a new secret.
Aggregated APIs
Everything your agent needs.One key.
5,000+ APIs, skills, and models behind one key — billed per call, no per-vendor contracts.
5,000+
APIs & Models
Up to 90%
off official API prices
One prompt
Get your AI quickly onto 5,000+ APIs.
Nothing to install and no key needed — paste it into the agent you already use, and it walks you from model choice to your first call, using our docs instead of guessing at endpoints.
Quickstart prompt
Please read the Felo API Platform documentation index below, then help me integrate the Felo API into my project: https://openapi.felo.ai/llms.txt About Felo API Platform: one key for chat and video models, an aggregated third-party API catalog, and Felo's own agent APIs. My question: I want to summarize a long PDF in my app — which model should I use, and how do I make my first call?
- 30 models, one key
- 5,000+ aggregated APIs, same key
- 9 first-party APIs with skills
- 3 wire protocols, same client
Works with the agents you already use
Every source you add costs you a key, a quota, and an invoice.
Adding a data source should be a line of code. In practice it is a vendor evaluation, a credential to store, a quota to watch, and a bill to reconcile.
An integration per vendor
Different auth, different pagination, different error shapes. The glue code is written from scratch every time, and it is yours to maintain.
No single view of spend
Cost is spread across invoices, so nobody can say what one agent run actually cost until the month closes.
5,000+ APIs and skills
Everything your agent can call through one Felo key.
Third-party services, called through the same key. No vendor account, no separate contract.
Similarweb
Traffic and engagement, ranking, traffic sources, audience, and keyword insights.
TikTok
Profiles, videos, comments, search, ads, and Shop data.
Instagram
Profiles, posts, comments, followers, reels, hashtags, and locations.
YouTube
Videos, channels, playlists, comments, trending, and regional search results.
X
Users, posts, lists, communities, trends, Spaces, and write actions.
Reddit
Posts, comments, subreddits, user profiles, and community activity.
Pinterest
Pins, boards, profiles, and public creative surfaces.
LinkedIn
Company pages, posts, and profile and company insights.
Telegram
Channel info, posts, and batch channel lookups.
Bilibili
Profiles, videos, comments, danmaku, search, and live rooms.
Perplexity
Web-grounded answers with citations, in light or stronger variants.
OpenAI Web Search
An OpenAI-model answer produced after a live web search, with URL citations.
Anthropic Web Search
A Claude-written answer produced after a live web search, with inline citations.
Tavily
Web search, page extraction, site crawling, and URL mapping.
Exa
Neural semantic search, page contents, and cited answers.
Firecrawl
Scrape pages to markdown, search, map URLs, and run async crawls.
CoinGecko
Live crypto prices, coin metadata, historical charts, exchanges, and trending.
Financial Data
Estimates, company facts, earnings, statements, ownership, rates, and filings.
Available routes change over time as the catalog grows.
Felo's own capabilities, not resold ones. Install the CLI once and all of them are available under the same key.
Install oncenpm install -g felo-ai
Search
Real-time web search and answer generation for agents, research tools, support bots, and content workflows.
PPT
Generate slide decks asynchronously from prompts, themes, and optional LiveDoc context, then poll task status and retrieve results.
Ready to use
SuperAgent
Create streaming agent conversations, consume SSE output, follow up inside the same thread, and attach LiveDoc context.
Image capable
Web Fetch
Extract webpage content as HTML, text, or Markdown with crawl mode, readability, selector, and timeout options.
Agent utility
X Search
Search X users, tweets, timelines, and replies with pagination metadata for research, monitoring, and content agents.
Limited-time free
Mindmap
Create structured mindmaps from a topic, question, or optional LiveDoc context for planning and research workflows.
Ready to use
Landing Page
Generate landing page content asynchronously, query task history, and retrieve generated HTML results for campaign workflows.
Ready to use
LiveDoc
Create knowledge bases, upload resources, import URLs, and run semantic retrieval for context-aware agents.
Knowledge workflow
YouTube Subtitles
Fetch YouTube subtitles by video code with optional language and timestamp parameters for media research agents.
Agent utility
Recently added LLMs
Newest first, by release date.
OpenAIGPT-6.1 Sol
An upgrade to GPT-6 Sol, still positioned below…
Context1.05M
Input tokens$2.00 /M
Output tokens$10.00 /M
AnthropicClaude Sonnet 5.5
Anthropic's Sonnet-class route for well-scoped…
Context1M
Input tokens$2.00 /M
Output tokens$10.00 /M
- OpenAI
GPT-6 Luna
The fast, cost-efficient tier of the GPT-6…
Context1.05M
Input tokens$0.10 /M
Output tokens$0.50 /M
xAIGrok 4.7
The flagship Grok route for coding, agentic…
Context500K
Input tokens$1.60 /M
Output tokens$4.80 /M
XiaomiMiMo V2.6 Pro
Xiaomi's flagship foundation model, built at a…
Context1M
Input tokens$0.43 /M
Output tokens$0.87 /M
DeepSeekDeepSeek V4.1 Flash
A sparse mixture-of-experts model from DeepSeek…
Context1M
Input tokens$0.15 /M
Output tokens$0.60 /M
- OpenAI
GPT-6 Astra
OpenAI's most capable model, built for the…
Context1.05M
Input tokens$10.00 /M
Output tokens$50.00 /M
Z.AIGLM 5.3 Flash
Z.AI's open-weight Flash route with a 1M-token…
Context1M
Input tokens$0.15 /M
Output tokens$0.50 /M
Models below official pricing. APIs billed per call.
Model routes are priced up to 50% below the official provider API rate. Third-party API calls are billed per call, from the same credit balance. Credits cost $0.001 each.
See pricingFinal cost depends on the model, input/output mix, cache usage, and active plan.
- No card required to sign up
- No subscription required — top up as you go
- One balance across models, APIs, and tools
- Usage logs and task history
Onboard your agent in under 1 minute.
One paste replaces the per-vendor setup: your console writes the prompt, your agent applies it, and nothing on your side changes.
- Example
Connect this project to the Felo API. Base URL https://openapi.felo.ai/api/v1 Models: GET /api/v1/models…
FELO_API_KEY=••••••••••••
From your consoleStep 1
Sign in and create a key
The console needs the key before it can write the prompt for you.
Connect this project to the Felo API. Base URL https://openapi.felo.ai/api/v1 Models: GET /api/v1/models…
⌘VTo your agent
Step 2
Copy the setup prompt
Your key is already inside it — paste it into your agent and it wires itself up.
- What is Reddit saying about this launch?What did Bitcoin do over the past day?Turn this pricing page into a JSON table.Make a slide deck out of these notes.
Step 3
Ask it something
Ask for a model or a catalog route. The first answer proves the route works too.
Questions, answered.
Short answers on how the aggregated endpoint works, what it costs, and what happens to your data.
What is the Felo aggregated API?
One endpoint and one API key that reach three things: an aggregated catalog of third-party APIs, frontier models from multiple providers, and Felo's own agent APIs. Requests are billed per call from a single credit balance, so you do not open a separate account with each vendor.
Can my agent call third-party APIs, or only models?
Both. The same key serves the model routes and the aggregated third-party API catalog, so an agent can call a model and then call a search, social, or market-data route without a second credential.
What is in the third-party API catalog?
Search, social media, go-to-market, business data, contact data, market intelligence, commerce, and prediction markets. The set of available routes changes over time as the catalog grows, so treat the categories as the shape of the catalog rather than a fixed list.
How is this different from integrating each provider directly?
You keep one account, one key, and one bill instead of a separate credential, quota, and invoice per vendor. Model IDs are stable, so moving between providers or between models from the same provider is a string change rather than a new integration.
Which wire protocols are supported?
Three: OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages. Existing clients for any of them work after a base URL and key change.
How does billing work?
Credits are consumed per request. Model usage and third-party API calls both draw on the same balance, and API calls are billed per call. Credits cost $0.001 each, and no subscription is required — top up as you go. Paid plans add a monthly credit budget. See the pricing page for the current rates.
Am I locked into specific models?
No. The catalog is discoverable at request time through the models endpoint, and any listed model ID works with the same key.
What happens if a provider has an outage?
Provider availability differs by route. Model discovery reports what is currently listed, and requests to an unavailable route return an error you can retry or route around.
What happens to my prompts and data?
Requests are processed to serve the response. Review Felo's privacy policy for retention and handling details before sending sensitive data.
Start with one key.
Create a key, point your client at the endpoint, and call your first model or API route today.
