Comparison

Bystep vs Conductor

A Mac app from Melty Labs (YC) that runs several coding agents side by side — Claude Code, Codex and Cursor — each in its own git worktree, with diff review, merge and GitHub/Linear integration. It also offers cloud workspaces. macOS only for now.

What Conductor is

A Mac app from Melty Labs (YC) that runs several coding agents side by side — Claude Code, Codex and Cursor — each in its own git worktree, with diff review, merge and GitHub/Linear integration. It also offers cloud workspaces. macOS only for now.

Pricing
Free · Pro $50 · Teams $60/user · Enterprise — as listed on conductor.build/pricing when we checked on 10 Sep 2026. Agents run on your own subscriptions or API keys. Check their page for current numbers.

Competitor facts checked 10 Sep 2026 · Sources: conductor.build/pricing, docs.conductor.build

Feature matrix

Where the two actually differ

Every Bystep item here is live in the product today — not a roadmap promise. Competitor claims we couldn't verify first-hand are marked “as far as we know.”

BystepConductor
PRD generationYesNoStarts from a workspace per task, not from an idea-to-PRD flow
Acceptance criteria per taskYesids like AC-checkout-1, derived from the PRD, editable by handNoNo structured acceptance-criteria field per task, as far as we know
Verification gateYes`task complete` returns 409 without passing evidence or a recorded overrideNoYou review the diff and merge yourself; nothing requires evidence before a task counts as done
Evidence panelYestest output, diff stat, files touched, optional screenshot — stored, not just claimedNoThe diff is the record — no stored proof page for a non-engineer
Works with any coding agentYesClaude Code, Cursor, Codex, Copilot, OpenCode, Gemini CLI via CLI + MCPYesClaude Code, Codex and Cursor on your own subscriptions or keys, locally and in its cloud workspaces
Import / export formatsYesSpec Kit, OpenSpec, Backlog.md, Task Master, Kiro, Claude Code Tasks, GitHub Issues, LinearNoIntegrates with GitHub and Linear, but is not a spec/task format hub
Client mode / proposalYesclient-viewer share page + proposal for the Studio tierNoBuilt for engineering teams (multiplayer); no client-facing progress or proposal page, as far as we know
Cost per featureYesper feature, not just per sessionNoNot shown per feature, as far as we know
Indonesian language & paymentsYesen/id UI, IDR pricing, QRIS/VANoEnglish product, USD pricing
Open-source CLIYes`@bystep/cli` + `@bystep/mcp`, MITNoClosed source

Honest answer

When to choose Conductor instead

  • You are an engineer on a Mac and want the most polished way to run several agents side by side, each in its own worktree, reviewing every diff yourself.
  • You want cloud workspaces and multiplayer for a dev team, with agents running on your own subscriptions or API keys.
  • You don’t need a PRD, acceptance criteria or stored evidence before a task counts as done — Conductor is built for parallel execution by engineers; Bystep is built to prove the result to people who don’t read diffs.

Keep your agents, add the proof

Bystep runs the agents you already use from one plan, then requires evidence before a task counts as done. The result is readable by people who don’t read diffs.