driftless vs drft
drft is a spec-driven governance layer for AI coding agents: it locks down requirements, plans, and approvals while agents like Claude Code or Cursor write the code. driftless is the full delivery platform: shaping, gating, shipping, tracking, and verification with persistent agents included. Every competitor fact on this page was verified against drft.build on September 24, 2026.
Verified against drft.build on September 24, 2026, including drft's own product FAQ. Competitor facts are re-verified before any change to this page ships.
Feature comparison
drft governs the agents you already run. driftless brings the agents and the governance together in one platform.
| Capability | driftless | drft |
|---|---|---|
| Product scope | All-in-one delivery platform: shaping, gating, shipping, tracking, and verification | Governance layer around external coding agents, by drft's own description |
| Coding agents | Persistent BA and coding agents included and orchestrated by the platform | No coding agents of its own; Claude Code, Cursor, Codex, Gemini, or any MCP agent does the coding |
| Spec lifecycle | Cascades: shape, gate, ship, verify with paired PRD and tech spec | Draft, refine, approve, plan, implement, verify with human approval gates at every stage |
| Planning | Orchestrated coding agents execute against context-engine directives: initiative details, task directives, acceptance criteria | Planning agents read your codebase via MCP and produce versioned implementation plans tied to the approved spec |
| Requirements shaping | BA Agent converts conviction into structured, testable, sequenced work | Refinement agents structure rough PM specs into versioned, machine-readable requirements |
| Audit trail | Comment threads as decision logs plus a full audit trail on tasks | Complete audit trail with actor, timestamp, and reason |
| Tracking and metrics | Kanban task board, nine dependency link types, delivery metrics, per-model cost dashboards | Team collaboration with role-based access |
| Spec-to-task traceability | Bidirectional: every requirement traces to a task, every task to a requirement | Versioned specs and plans tied to the approved spec |
Pricing comparison
Both approaches leave model usage on your provider bill. The difference is what ships today: driftless has paid tiers in production, drft lists a free tier with a Pro tier coming soon.
| Pricing | driftless | drft |
|---|---|---|
| Pricing model | Per seat, flat monthly price | Free tier today; a Pro tier is listed as coming soon |
| Free tier | $0: 1 user, 2 agents (BA + Coding), BYOK inference | $0: 1 workspace, 1 project, 3 team members, unlimited specs |
| Paid tiers | Pro: $100/seat/mo for the first 100 seats, $70/seat after. Enterprise: custom, per seat | Pro tier coming soon, per drft.build |
| Model cost | Billed by your provider under BYOK | Billed to your existing provider; drft never proxies or meters AI tokens |
| Platform scope included | Agents, task board, spec gates, analytics, and cost tracking in the platform | Governance of specs, plans, and approvals for agents you already run |
FAQ
- How is driftless different from drft? drft describes itself as a governance layer: it governs requirements, plans, and approvals while external coding agents such as Claude Code or Cursor do the coding through MCP. driftless is a delivery platform with its own persistent agents: a BA agent shapes paired PRDs and tech specs, the orchestrator runs coding agents against task directives, and the platform verifies acceptance criteria. With driftless you get shaping, execution, tracking, and verification in one place. With drft you add governance on top of agents you already run.
- How does pricing differ between driftless and drft? drft lists a free tier at $0 with 1 workspace, 1 project, 3 team members, and unlimited specs, and states that a Pro tier is coming soon. driftless ships paid tiers today: a free tier with 2 agents and BYOK inference, Pro at $100 per seat per month for the first 100 seats and $70 per seat after, and custom Enterprise plans. Both approaches bill model usage to your own provider: driftless through BYOK, drft by never proxying or metering AI tokens.
- Does driftless run the agents itself? Yes. driftless includes a BA agent for shaping specs and coding agents that the orchestrator directs against task directives, with delivery metrics and per-model cost tracking. drft connects to coding agents you already run, such as Claude Code, Cursor, Codex, or Gemini, through MCP.
See the difference in your own workflow
Start free with two agents and BYOK, or read the pricing page for tier details. The Cascades page explains the methodology the platform enforces.