Replit published Replit Missions on its official blog on November 17 and promoted it again from its official X account on November 19. You define the goal, the spending limit and which actions need your approval, and Replit coordinates a swarm of agents to plan, build, test and iterate in parallel. An independent AI judge checks results against the agreed success criteria; a mission completes only when they are met and stops if the budget runs out. Replit used it to build DataCat, its own application monitoring platform, in 2 days within a $3,000 budget, and in its NumPy tests a 32-agent swarm reached speedups of up to 19x.
Key Takeaways
- ✓How it works: you set the goal, a spending cap and which actions need approval (such as contacting customers, publishing or raising the budget), and an agent swarm works in parallel
- ✓Completion: an independent AI judge checks results against the agreed success criteria; if the budget runs out first, work stops with the outcome unresolved
- ✓Case study: DataCat, a monitoring platform built in 2 days within $3,000; it alerted on a deliberately injected database bug in 38 seconds and passed all 51 data and alert checks
- ✓Replit NumPy tests across 13 tasks: a single Auto (Max) agent reached up to 4x speedups, a 16-agent swarm 17x and a 32-agent swarm 19x
- ✓Replit comparison: from the same prompt, DataCat ingested 49,000 traces per second, about 5x the Claude Code /goal build; pricing was not stated in the blog post

Key Decision Metrics at a Glance
Turn your technical choice into a development budget
Compare 40 dev plans & simulate token costs vs $20/mo subscriptions
Project Links & Resources
Direct AccessIn-Depth Technical Analysis
Background
A single coding agent can already work for hours, but goals like "win our first 10 enterprise customers" or "replace a legacy system" need several parallel attempts plus someone managing budget and risk. Replit says its own team never had the capacity for this kind of work.
How it works
Missions first asks questions to define success, then records it in a Mission agreement: the spending limit, which actions need human approval and which tradeoffs are acceptable. Replit then coordinates a swarm of cloud agents that research, build and test in parallel, share findings and report progress at the frequency you choose. An independent AI judge checks results against the criteria and the agents refine their work from its feedback. A mission completes when the criteria are met and stops if the budget runs out.
Benchmarks
All figures are Replit's own tests. On 13 NumPy optimization tasks, a single Auto (Max) agent reached up to 4x speedups, a 16-agent swarm 17x and a 32-agent swarm 19x. DataCat was built in 2 days within $3,000 and alerted on an injected database bug in 38 seconds. From the same prompt it ingested traces about 5x faster than a Claude Code /goal build, though that build covered more product areas.
What it means for developers
Start a mission at replit.com/missions, write measurable success criteria and a budget cap first, and require approval for contacting customers, publishing and raising the budget. The blog post does not list pricing, so check what your account shows.
Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.