RXed AI News

AI to the bone.
Audited 2026-07-24 · RXed table v1.0

Gumloop

Visit gumloop.com
“Build AI agents for work” — the vendor’s own words

The no-code agent builder non-engineers actually keep using: every model included, deep MCP, a real audit trail. You price your workflows, not your seats, and the credit meter is where the surprises live.

The verdict, read by Jan · AI clone of Jan’s voice, generated locally with his consent.
Best for: Marketing, sales, ops and support teams that want to automate AI-heavy, multi-step work without waiting on engineering — and are willing to watch the credit meter as workflows scale.
Scope15/20
Quality7/10
Where the quality sits
8Reactive
Retrieval & Memory
7Orchestration
7Validation
7Models
SMB toolAutomation & AgentsProductivityResearchFreemium
Vendor
Gumloop, Inc. (formerly AgentHub) · www.gumloop.com
Origin
US — San Francisco
Pricing
Pro $37/mo · Enterprise Custom
Users (official only)
Not publicly disclosed (private WAU chart shown, no figure); ~24 employees as of Mar 2026. Named customers: Shopify, Ramp, Gusto, Samsara, Instacart, Opendoor. (source, 2026-03-12)
Pro$37/mostarts at; 20k+ credits, unlimited seats
EnterpriseCustom

Credit-based, not unlimited: Pro starts at $37/mo for 20,000+ credits and scales with credit volume, 20% off on annual billing, 14-day free trial and no permanent free tier.

checked 2026-07-25 · vendor pricing page

Element scores

Reactive
Retrieval & Memory
Orchestration
Validation
Models
Primitives
Compositions
Deployment
Emerging
Tap or hover any element to see why it got that score.

Strengths

Non-engineers pick it up in minutes and keep using it — that stickiness is what won Benchmark's $50M Series B in March 2026. Model-agnostic with every major provider bundled, 150+ connectors, and one of the deepest MCP implementations in the category (MCP nodes, a 40-tool MCP server, custom MCP servers). Gumstack gives security teams a single audit trail over every agent tool call across the org, and it is SOC 2 Type 2 and GDPR compliant.

Honest dings

Credit-based pricing is the real cost question: advanced-model calls burn ~10x the credits of standard ones, and heavy AI usage can push a team well past the $37 Pro headline. Embeddings, a hosted vector store, fine-tuning and synthetic data are simply not there — this is an orchestration layer, not an AI platform. Evaluations are listed but thin, and multi-agent coordination is still young.

Prices and details change — this passport is re-verified at least quarterly.
Sources (9) — every claim traceable

Every audit lists the research it rests on — transparency and traceability are the product. Tools evolve: each audit is a snapshot of its audit date, and re-audits supersede older versions (kept below for reference).