Runtime as featured inForbesRead the article

Hyperagent vs Convey

Hyperagent vs Convey: a self-serve platform for teams of AI agents compared with digital teammates trained by screen share. Features, security, deployment, pricing, and where Runtime fits for fintech.

Updated October 7, 20265 min read

TL;DR: Hyperagent suits teams that want a flexible, self-serve agent platform with a wide model catalog and public credit pricing. Convey suits back-office teams that want to train teammates by demonstration, including in legacy systems, with SOC 2, HIPAA, and a guided 30-day onboarding. For payment operations on regulated data in your own cloud, Runtime is the third option.

Feature
HyperagentHyperagent
ConveyConvey
RuntimeRuntime
What it isPlatform for teams of AI agentsDigital teammates for enterprise opsAgent harness for payment and fintech teams
How you buildConfigure agents, skills, memoriesDescribe it or share your screenAgents from your SOPs
Typical workResearch, marketing, ops, softwareInvoices, orders, reconciliation, ad opsPayment ops, risk, recon, support
Legacy UIsBrowser via BrowserbaseUI automation, a core strengthBrowser on the agent's computer
Where agents runHosted, private computer per threadHosted, tenant-specific sandboxesYour AWS, GCP, Azure, or self-hosted
ModelsWide hosted catalogVendor-managed, zero data retentionAny model, including open-weight
Published complianceNot publishedSOC 2 Type II, HIPAAAudit trail exportable for examiners
Getting startedSelf-serve30-day guided onboardingPayments-native forward-deployed engineer
PricingCredit plans from $20/monthNot publishedFree, Teams from $99/seat

Hyperagent and Convey are two of the most visible 2026 platforms for AI teammates that do real operational work. Both let non-engineers put agents to work across a company, and both talk about agents as colleagues rather than chatbots. They differ on how you build, what kind of work they target, and how you buy.

Hyperagent vs Convey at a glance

Hyperagent started inside Airtable and was spun out as an independent company in 2026, led by Airtable co-founder Howie Liu. You build teams of specialist agents with their own identities, instructions, tools, skills, and memories. Agents delegate to each other, run on schedules or watch for changes, and are called from Slack, Telegram, email, webhooks, or MCP clients. Each thread gets a private computer to write and run code, plus a browser, and outputs such as webpages, slides, and documents land in a shared library.

Convey builds digital teammates that operators train by describing a process or sharing their screen. Each teammate has its own identity and role-based access to apps, works through APIs and by clicking through software like a person, and flags a human when something is off. According to a16z, which led Convey's $38M Series A in June 2026, it compiles what it learns into versioned, testable programs. Published customers include NBCUniversal, TelevisaUnivision, Samsara, Unity, Faire, and ChargePoint.

Product scope

Hyperagent. Broad by design: research, marketing, recruiting, software, operations, creative production, and business intelligence. Its strength is flexibility, with rich outputs, delegation between agents, and rubrics to evaluate quality over time.

Convey. Focused on recurring back-office workflows: invoice processing, order entry, reporting, campaign operations, and month-end reconciliation across finance, accounting, ad ops, HR, and more. Its strength is turning what an operator already does on screen into dependable automation.

Security and deployment

Hyperagent. A hosted platform with workspaces, roles, an Ask first approval mode, approval cards that preview exact writes, and a full transcript per thread. Its docs note that Ask first is a behavioral preference, not a sandbox. Certifications and a self-hosted option are not published in its public docs.

Convey. A hosted platform with a published enterprise posture: SOC 2 Type II, HIPAA with BAAs for qualifying customers, regular penetration testing, SAML SSO and MFA, isolated agent identities, audit logs, tenant-specific sandboxed execution environments, and zero data retention with model providers.

Models

Hyperagent. One of the widest hosted model catalogs available, including Anthropic, OpenAI, Google, xAI, DeepSeek, Kimi, Qwen, and GLM, plus the option to run GPT usage on a connected ChatGPT subscription. You pick the model per agent and can cap spend per run.

Convey. Models are managed by Convey, under zero data retention agreements with its providers. Model choice is not published.

Implementation and support

Hyperagent. Self-serve. You sign up, configure agents, and learn the platform yourself, with detailed public documentation.

Convey. Sold through sales with a 30-day onboarding, "forward deployed" IT specialists who configure the first teammates, and dedicated solutions engineers.

Pricing

Hyperagent publishes credit plans from $20 to $10,000 per month. Your plan price becomes credit, plus a bonus of up to 45 percent on larger plans, spent on models, browser sessions, search, and tools.

Convey: Pricing is not published; it is sold through sales, with plans that Convey says scale with the value you get.

Where Runtime fits

Hyperagent and Convey are horizontal: they aim to serve every department. Payment and fintech operations need three things neither publishes.

Domain depth. Payment work is mostly exceptions: a held payout, an ACH return with an odd code, a dispute where the evidence contradicts the claim. Runtime agents start read-only, and approvals sit in front of releasing a payout, applying a reserve, booking a ledger entry, or replying to a sponsor bank. Every run is recorded end to end, and the record your sponsor bank asks for stays with you.

Your cloud, your sandbox, your models. Runtime runs agent computers in your AWS, GCP, or Azure account, or fully self-hosted, with your own sandbox provider and any model, including open-weight models served in your cloud for PCI and PII work. Card numbers and SSNs are stripped from prompts and logs.

A payments-native engineer. Runtime's forward-deployed AI engineers built payment and fintech infrastructure at Hulu's payments team at Disney, Finix, Modern Treasury, and BlackRock's AI quant group. They map your processes and build the first agents with your team. Once a process is solved, agents can turn it into a deterministic script that lives in your repos, so engineers own it if it becomes mission-critical.

Runtime also connects the teams. A single stuck payment touches support, payment ops, finance, and risk; on Runtime it is one investigation with one record.

Which should you choose

  • Choose Hyperagent if you want a flexible, self-serve agent platform across many functions, a wide model catalog, and public credit pricing.
  • Choose Convey if your work lives in legacy UIs, operators want to teach by screen share, and SOC 2 and HIPAA on a hosted platform cover your requirements.
  • Choose Runtime if your agents work card data, PII, ledgers, or payouts and must run in your cloud on your models, with approvals before money moves and an engineer who knows payments.

See Runtime on your busiest queue

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

Frequently asked questions

What is the difference between Hyperagent and Convey?

Hyperagent is a self-serve platform where you configure teams of agents with identities, skills, memories, and a private computer per thread, called from Slack, email, Telegram, webhooks, or schedules. Convey builds digital teammates that operators train by describing a process or sharing their screen, aimed at back-office work across departments, and sold with a guided onboarding.

Which is better for legacy systems?

Convey is built for it: teammates work through APIs and by clicking through software like a person, and operators teach them by screen share. Hyperagent agents can drive a browser too, but its focus is broader research, building, and delivery work.

Is Hyperagent or Convey SOC 2 compliant?

Convey says it is SOC 2 Type II certified and HIPAA compliant, with SAML SSO, audit logs, and zero data retention with model providers. Hyperagent's public documentation covers roles, approvals, and data deletion but does not publish its certifications.

Can either run in my own cloud?

Neither publishes a bring-your-own-cloud or self-hosted option. Runtime runs agent computers in your AWS, GCP, or Azure account, or fully self-hosted, with your own sandbox provider and models.

Related comparisons