AI Potluck
Product / UX / Orchestration & agents

Replit Agent

Replit

Cloud development agent embedded in Replit's online IDE that builds applications from a natural-language description, scaffolding projects, installing packages, writing code, configuring databases and deploying to production. It runs entirely in the hosted environment, with plan mode, agent skills, automations, a task board and end-to-end publishing.

replit.com answers 403 to an automated fetch, so the record rests on Replit's own documentation. Verified 2026-08-13 via the Replit Agent documentation.

Openness

1 high confidence
1.0
source
closed
license
proprietary(Replit cloud SaaS)
self-host
no

Proprietary cloud agent embedded in Replit's hosted IDE; no source or self-hosting. The product page is no longer fetchable - replit.com answers 403 to a non-browser request - so the claim rests on Replit's own documentation, which describes the agent entirely inside the hosted platform (deployments, machines, hosted database, Replit Auth and SSO, workspace security, monetization) with no source distribution and no self-host path.

  • https://replit.com/agent recorded 2026-06-04

    Replit Agent proprietary AI app-building product within Replit's cloud platform. Not re-fetchable - replit.com answers 403 to a non-browser request.

  • https://docs.replit.com/replitai/agent recorded 2026-08-13

    Replit Agent documented wholly inside the hosted platform - plan mode, agent modes, publishing and deployments, machines, hosted database, Replit Auth and SSO, security centre, monetization; no source download and no self-host option

Adoption

4 medium confidence
4.0

Replit reported more than 50 million platform users, with agent-driven consumption as the primary growth driver. That count is platform-wide rather than agent-specific, so the level sits conservatively at 4 on reported traction rather than where 50M would otherwise land.

  • https://sacra.com/c/replit/ recorded 2026-08-13

    Replit user base 22.5M in April 2023, 40M+ by September 2025 and 50M+ by March 2026, users from 85% of the Fortune 500, growth attributed to usage-based agent pricing

Capability

3 low confidence
3.0

Capable autonomous app-builder but oriented to full-app generation/deploy rather than benchmark-graded SWE tasks; no first-party benchmark, scored 3 on feature matrix.

  • https://docs.replit.com/replitai/agent recorded 2026-08-13

    plan mode, agent modes, model selector, app testing, publishing and deployments, agent skills, automations, web search, task board and artifact types; no benchmark figure published

Verified 2026-08-13