Runtime as featured inForbesRead the article

Runtime vs Dust

Runtime vs Dust for fintech and payment teams: a horizontal platform for company-wide knowledge agents vs an agent harness for payment operations in your cloud, with approvals on money movement.

Gus Trigos
Co-founder and CEO, Runtime
Updated October 8, 2026 · 7 min read

TL;DR

Dust is a strong horizontal platform for agents that use company knowledge across sales, support, engineering, legal, and people teams, with solid governance and a wide choice of models. Runtime is the stronger choice when the agents do mission-critical payment operations work: on their own computers in your cloud, with approvals before money moves and an audit trail for your sponsor bank.

Feature
RuntimeRuntime
DustDust
What it isRuntimeAgent harness for payment operationsDustMultiplayer platform for company agents
Best atRuntimePayment exceptions, recon, risk, disputesDustKnowledge work across every department
ConnectorsRuntimeAPIs, databases, CLIs, MCP, browserDust70+ tools plus MCP
Where agents runRuntimeOwn computers in your cloud, or self-hostedDustDust cloud, US or EU; single-tenant on Enterprise
ModelsRuntimeAny, including open-weight in your cloudDust20+ frontier models
Approvals on money movementRuntimeBuilt inDustHuman review points
Audit trailRuntimeEvery query, tool call, approval, costDustAudit logs on Enterprise
Getting startedRuntimeForward-deployed AI engineerDustSelf-serve, CSM on Enterprise
PricingRuntimeBy request; free pilot for approved teamsDustFree, Pro $24, Max $120 per seat

Dust is the better choice for company-wide knowledge agents that answer, draft, and triage from more than 70 connected tools across every department. Runtime is the better choice when agents do payment operations work on regulated data, in your cloud, with approvals before money moves. Runtime is the AI agent harness for payment and fintech teams.

Dust and Runtime at a glance

Dust is a platform for human-agent collaboration. Teams build agents with instructions, knowledge, skills, and tools, connect them to more than 70 systems (Slack, Notion, Google Drive, GitHub, Salesforce, Zendesk, and any tool through MCP), and choose from more than 20 models from OpenAI, Anthropic, Google, Mistral, and others. Agents run on demand, on a schedule, or from webhooks such as a new Zendesk ticket or GitHub pull request. Dust positions itself for "AI Operators," the people who build agents for their teams, and says 3,000+ organizations including Datadog, 1Password, Vanta, and Clay have deployed 300,000+ agents across sales, support, engineering, legal, and people teams.

Runtime is the AI agent harness for payment and fintech teams. Payment ops, risk, compliance, finance, underwriting, onboarding, and support build agents from their SOPs. Each agent works on its own isolated computer in your cloud, reaches your ledger, processor, data warehouse, and ticketing tools through APIs, databases, CLIs, and MCP servers, and uses a browser for portals with no API. Agents start read-only, stop for approval before anything moves money, and record every run end to end.

Who each is for. Dust fits companies that want one place for every department to build knowledge and productivity agents. Runtime fits payment companies that want agents doing the operational work on money, under the controls a sponsor bank expects.

How they differ

Company knowledge vs payment operations

Dust is horizontal by design, and its strength is company context: an agent that answers from the wiki, drafts a sales follow-up from CRM notes, or triages a support ticket using past tickets. Payment operations need a different kind of agent. Tracing a stuck payout means querying the ledger, matching the processor record, reading the bank file and return codes, and sometimes logging into a sponsor bank or network portal with no API. Runtime agents do that on their own computers, with a shell, a browser, and scoped credentials, and they reason through the exception rather than retrieve an answer.

Where agents run and who holds the data

Dust runs as a managed cloud service with US or EU data residency, and Enterprise adds single-tenant deployment. Dust doesn't train on your data and has zero data retention agreements with model providers. Runtime puts agent computers in your own AWS, GCP, or Azure account, or you self-host the whole platform with Helm. Credentials are masked, card numbers and SSNs are stripped from prompts and logs, and PCI data can stay in your environment.

Models

Dust gives you a broad menu of frontier models and lets each agent use the right one. Runtime also lets you bring any model, and adds open-weight models served inside your cloud for PCI and PII work, plus routing across harnesses (Claude Code, Codex, OpenCode) and models with fallbacks when a provider goes down.

Controls on money movement and the audit trail

Dust has a solid permission model (spaces, user/builder/admin roles, agent access separated from user access) and human review points in workflows, with audit logs on Enterprise. Runtime's guardrails are specific to money: agents start read-only, and you choose the steps that need a person, such as releasing a held payout, applying a reserve, booking a ledger entry, or replying to a sponsor bank. Every run is recorded with the trigger, every query and tool call, approvals, cost, and result, so the record your sponsor bank asks for stays with you.

Self-serve vs forward-deployed engineer

Dust is self-serve on its Business plan, with a customer success manager on Enterprise. Runtime pairs you with a forward-deployed AI engineer whose team built payment infrastructure at Hulu's payments team at Disney, Finix, and Modern Treasury. The FDE maps your processes, builds the first agents with your team, sets up guardrails, and trains your admins. Once a process is solved, Runtime agents can turn it into deterministic code that lives in your repos, so engineers can own it if it becomes mission-critical.

Where Runtime enters the picture

Dust gives every department agents that know the company. What it leaves open for a payment company is the operational work itself. A merchant asks in Slack why a payout is held, and the answer is not in the wiki. An agent has to query the ledger, match the processor record, read the bank file, check the risk notes, and draft a release that a person approves before any money moves. That needs an agent with its own computer inside your environment, not a knowledge assistant.

Runtime is the AI agent harness for payment and fintech teams. It runs that work in your cloud, and it sits comfortably next to Dust: Dust for company knowledge, Runtime for the queues that touch money.

  • No single vendor to depend on. Dust offers a wide choice of hosted models. Runtime also lets you choose the cloud and the sandbox: agent computers run in your AWS, GCP, or Azure account or fully self-hosted, sandboxes fail over, and models span Claude, GPT, Gemini, and open-weight, with harness routing across Claude Code, Codex, and OpenCode.
  • Guardrails you set. Agents start read-only, releasing a payout or booking a ledger entry waits for a named approver, RBAC, masked credentials, card numbers and SSNs stripped from prompts and logs, and every run recorded.
  • Built for teams. Support, payment ops, finance, risk, and compliance share agents, templates, and one memory organized around the transaction, not the document.
  • An engineer, not an account executive. A forward-deployed AI engineer whose team built payment infrastructure at Hulu's payments team at Disney, Finix, Modern Treasury, and BlackRock maps your processes and builds the first agents with your team.
New agent

Create a banking ops agent that handles RFIs, hold-harmless requests, and returns of funds from our partner bank, verifies against our ledger, and drafts replies for my approval.

Describe the agent you want
Agent ready

bank-ops-agent

Tools picked from your prompt

EmailLedgerSFTPMastercard
Describe the work in plain English, attach the SOP, and Runtime builds the agent.

Where Dust is stronger

  • Breadth across the company. Sales, marketing, legal, engineering, and people teams all get agents on one platform.
  • Out-of-the-box connectors. 70+ integrations for knowledge sources, so agents have company context on day one.
  • Self-serve speed. Builders can create and share agents without an implementation project.
  • Transparent, low entry price. Published seat pricing with a free tier.
  • Proven adoption. Thousands of organizations and a large library of agent patterns.

Pricing

Dust: The Business plan (up to 100 people) has three seat types: Free ($0, 500 credits one time), Pro ($24 per seat per month billed yearly or $30 monthly, 8,000 credits per seat per month) and Max ($120 per seat per month billed yearly or $150 monthly, 40,000 credits). Programmatic usage is $0.01 per credit. Enterprise is custom and adds pooled credits, SCIM, audit logs, custom retention, and single-tenant deployment.

Runtime is available by request, with custom pricing and self-hosting. Approved teams pilot it free on one real workflow first. Teams usually weigh it against the work of analysts they were about to hire. Rain saved $750k in vendor and salary spend with Runtime, which it uses across technical payment support, fraud and risk transaction monitoring, compliance, cybersecurity, payment operations, engineering investigations, PR reviews, QA testing, and reporting.

Which should you choose

Choose Dust if

  • You want company-wide knowledge and productivity agents for every department
  • Most of the value is in answering, drafting, and triaging from company context
  • Self-serve building with published seat pricing matters most
  • Agents won't need to touch payment systems or move money

Choose Runtime if

  • The work is payment exceptions, reconciliation, disputes, alerts, or bank RFIs
  • Agents need your ledger, processor, data warehouse, and portals with no API
  • Card data or PII must stay in your own cloud, with models you choose
  • You need approvals on money movement and an audit trail your sponsor bank will accept

Some fintechs run both: Dust for company knowledge, Runtime for the operations that touch money.

Frequently asked questions

What is Dust?

Dust is a platform where teams build AI agents that use company knowledge and tools. It connects to more than 70 tools, including Slack, Notion, Google Drive, GitHub, Salesforce, and Zendesk, supports more than 20 models, and lets agents run on schedules or webhooks. Dust says 3,000+ organizations use it.

Is Dust secure enough for a fintech?

Dust is SOC 2 Type II certified and GDPR compliant, offers US and EU data residency, does not train models on customer data, and has zero data retention with model providers. Enterprise adds SCIM, audit logs, and single-tenant deployment. If you need agent computers inside your own cloud account or a fully self-hosted platform, Runtime offers both.

Can Dust agents move money or update a ledger?

Dust agents can call tools through connectors and MCP, so technically they can take actions. Runtime is built for that work specifically: agents start read-only, and steps like releasing a payout, applying a reserve, or booking a ledger entry wait for a named approver, with the full run recorded.

How much does Dust cost?

Dust's Business plan, for up to 100 people, has Free ($0, 500 credits), Pro ($24 per seat per month billed yearly, $30 monthly, 8,000 credits) and Max ($120 per seat per month billed yearly, $150 monthly, 40,000 credits) seats. Enterprise is custom.

Can we run Dust and Runtime together?

Yes. Dust can serve company-wide knowledge and productivity agents while Runtime runs the payment operations agents that touch the ledger, processor, and bank. They connect to many of the same tools, such as Slack and Salesforce.

Where does Runtime fit next to Dust?

Dust is a horizontal platform for knowledge and productivity agents across a company. Runtime is the AI agent harness for payment and fintech teams: agents investigate held payouts, recon breaks, disputes, and alerts on their own computers in your cloud, with approvals before money moves. Many fintechs run both.

See Runtime on your most manual payment workflow

Bring one SOP. A forward-deployed AI engineer builds the first agent with your team, inside your cloud.