Autobound

API and MCP server delivering 700+ B2B signals across 35 signal types from 35+ public sources, sold as one-time credit packs instead of a monthly seat.

RecommendedLast reviewed September 17, 2026

Get Autobound autobound.ai ↗

Best for

AI-nativeB2B SaaS

Ease of use

Medium

Getting a signal into a workflow means wiring the API or MCP server and tracking credit consumption yourself; there's no click-through dashboard sitting between a signup and a working feed.

Interface

APIAgent-ready

Pricing

One-time credit packs, not a subscription. Starter is $19 for 2,000 credits; Enterprise is $4,999 for 1,249,750 credits. Credits never expire, and signup gives 1,000 free credits with no card required.

Need it run for you?

We design and run stacks like this on client pipelines.

How we classify tools

In our stack means we run it on live client work; Recommended means we evaluated it and would point a client to it, without hands-on usage claims; Previously used means we migrated away and say why.

Best-for chips are capped at two per tool. Ease is judged for the tool’s majority buyer: easy is productive in an afternoon, medium takes about a week of configuration, hard needs an operator. Agent-ready means it runs headlessly via API, CLI, MCP, or webhooks.

No star ratings anywhere: scores imply a measurement nobody publishes, so we give verdicts in sentences instead. Reviews are dated, pricing is verified against the vendor’s page as of that date, and paid links are disclosed where they appear.

The full methodology

About

What Autobound actually is.

Autobound calls itself the signal layer for enterprise GTM. It pulls 700+ signals across 35 signal types from 35+ public sources, covering 250 million-plus contacts and 50 million-plus companies. Access runs through a REST API answering in under 200 milliseconds, a native MCP server, scheduled file exports to cloud storage, or white-label OEM licensing. There's no dashboard in that list. Autobound builds for a team wiring a signal feed into something it already runs, a Clay table, a Claude Code agent, an internal alert system, not for someone who wants to log in and browse.

Pricing follows the same logic. Credits get bought once and never expire, so a team pays for what it queries instead of a seat count sitting idle: 2 credits per Search result, 2 per Enrich, 1 per Buyer Intent contact. That's useful for usage that spikes unevenly, but it also means engineering time goes in before anything shows up, and the bill moves with query volume, not a flat number.

What it does

Tracks 700+ signals across 35 signal types, including executive hires, hiring surges, promotions, funding rounds, and technology adoption.
Pulls from 35+ sources: SEC filings, LinkedIn, job boards, patent databases, Glassdoor, Reddit, GitHub, news feeds, and earnings call transcripts.
Covers 250 million-plus contacts and 50 million-plus companies, with 99%+ domain coverage and 90%+ LinkedIn URL matching for entity resolution.
Responds through a REST API in under 200 milliseconds, or ships as scheduled batch files to Google Cloud Storage or S3.
Ships a native MCP server plus OpenAI function calling, so an agent queries signals without a custom integration layer.
Runs signals through a six-stage pipeline: ingestion, extraction, normalization, enrichment, entity resolution, and delivery.
Refreshes data daily and keeps a timestamped audit trail back to the original source.

None of that shows up in a dashboard. It shows up in an API response.

In the system

Where it sits in the pipeline.

Runs well with

The tools we'd run next to Autobound, and what each pairing covers.

ClayAutobound's own customer list names Clay; a table already running enrichment waterfalls can pull signal data in through the same API call.

Claude CodeAutobound documents Claude Code workflows directly, so a signal query runs inside an existing agent build instead of a separate integration.

SmartLeadA signal the API surfaces (a funding round, an exec hire) can trigger a sequence the moment the call returns, instead of sitting in a report.

What it costs

Pricing, without the fiction.

From the Autobound pricing page · verified September 17, 2026

Starter

$19 one-time

one-time credit pack, no recurring charge

2,000 credits ($0.0095/credit)
Full REST API + MCP server access to 35+ signal sources
Vendor labels this tier for testing and proof-of-concept work

Growth

$49 one-time

one-time credit pack, no recurring charge

5,444 credits ($0.009/credit)
Same API and MCP server access as Starter
Vendor labels this tier for production workloads

Scale

$149 one-time

one-time credit pack, no recurring charge

19,867 credits ($0.0075/credit)
Same API and MCP server access as lower tiers
Vendor labels this tier for high-volume teams

Pro

$499 one-time

one-time credit pack, no recurring charge

83,167 credits ($0.006/credit)
Same API and MCP server access as lower tiers
Vendor labels this tier for enterprise integrations

Business

$1,299 one-time

one-time credit pack, no recurring charge

288,667 credits ($0.0045/credit)
Same API and MCP server access as lower tiers
Vendor labels this tier for enterprise scale

Every tier buys the same 35+ sources and 700+ signals across 35 signal types through the same REST API and MCP server; the only thing that changes between tiers is credit volume and the per-credit rate, which drops from $0.0095 at Starter down to $0.004 at Enterprise. Consumption is metered per action rather than per seat: a Search costs 2 credits per result, an Enrich costs 2 credits per signal, Buyer Intent runs 1 credit per contact, a Bulk Export runs 50 credits plus 5 per topic, and a zero-result query costs nothing. That model rewards uneven, spiky usage over a flat monthly seat, but it also means the bill scales with how much a build actually queries, not a number fixed at signup. Teams that outgrow the $4,999 Enterprise pack can move to custom volume and rates past 11.5M credits.

The verdict

Choose it, or skip it.

Choose it if

A team that wants to build its own signal-triggered workflow (an agent, a Clay table, an internal alert) and would rather call an API than learn a new dashboard.
AI-native builds: the native MCP server, OpenAI function calling, and documented Claude Code workflows mean an agent can query signals with no custom integration layer.
Usage that spikes unevenly rather than running at a flat volume every month, since credits are bought once, never expire, and only draw down on an actual Search, Enrich, or Buyer Intent call.
Anyone who wants to try the data before paying: 1,000 free credits on signup, no card required, against the same 35+ sources every paid tier uses.

Skip it if

A team that wants a dashboard to browse accounts and signals in a UI; Autobound documents a REST API, an MCP server, file delivery, and OEM licensing, but no end-user application.
Anyone who wants a flat, predictable monthly bill; credit consumption is metered per Search, Enrich, or Buyer Intent call, so cost moves with query volume, not a fixed seat price.
A buyer who needs the vendor to do the integration work; getting a signal into a live workflow is on the team wiring in the API or MCP server.

Alternatives

If not Autobound, then what.

6senseContact-sales only. The pricing page names three Sales Intelligence bundles (Predictive AI, Data Credits, or both combined) but shows no dollar figure anywhere on the site.
Pick this when the buying committee wants a dashboarded ABM platform with account scoring built in, not a raw signal feed to build a UI around.
WarmlyNo published free tier; paid plans start at $10,000/yr ($4,875/quarter) for web deanonymization, scaling to $20,000 to $30,000/yr for the AI chat and autopilot tiers.
Pick this for identified-visitor and buying-signal orchestration with a workflow builder and interface included, instead of assembling that layer on top of a credit-metered API.
RB2BFree tier (150 resolutions/mo); paid plans from $79/mo (Starter), Pro from $149/mo
Pick this for a simpler, interface-first website-visitor identification tool when the need is one signal type, not 700+ across 35 sources.

FAQ

Asked in real client work.

How much does Autobound cost?

It sells one-time credit packs, not a subscription. Starter is $19 for 2,000 credits and Enterprise is $4,999 for 1,249,750 credits, with Growth, Scale, Pro, and Business in between. Credits never expire.

Is there a free trial?

Yes. Signing up gives 1,000 free credits with no card required, enough to test the API and MCP server against the same 35+ sources every paid tier uses.

What actually uses up a credit?

A Search costs 2 credits per result, an Enrich costs 2 credits per signal, and Buyer Intent costs 1 credit per contact. A Bulk Export runs 50 credits plus 5 per topic, and a zero-result query is free.

Can an AI agent query Autobound directly?

Yes. It ships a native MCP server for Claude, Cursor, and other MCP clients, plus OpenAI function calling and documented Claude Code workflows.

Does Autobound have a dashboard?

Not one the vendor documents. Access is a REST API, a native MCP server, scheduled file exports to cloud storage, or white-label OEM licensing, not a login-and-browse application.

What happens past the largest published tier?

Autobound offers custom volume and rates beyond 11.5M credits for teams that outgrow the $4,999 Enterprise pack.

Ready to engineer your GTM motion?

Tell us how your motion runs today. We'll show you what we'd engineer.