FAQ
Pricing, limits, common questions.
Pricing
- Free: 1 workspace, 100 tasks/month, GitHub only, Claude only. No credit card.
- Team: $499/workspace/month. Unlimited tasks. GitLab + Codex unlocked.
- Self-host: contact us for a quote. Includes SSO, audit export, support SLA.
No per-token resale. You pay vemlor for the platform; you pay Anthropic/OpenAI directly for inference.
Why not credits?
Credits make pricing predictable for vendors and unpredictable for customers. We took the opposite trade. Flat per-workspace fee, you manage your own LLM bill.
What if I exhaust my LLM subscription?
The agent fails the run with a clean error. Your workspace stays healthy; only that one task is affected.
Is self-hosting really free?
The binary is. Support and SLAs are paid. SSO and SAML are gated to Self-host paid tier.
How do you handle prompt injection?
Runner pods are network-restricted (egress to git host + LLM API only) and ephemeral (destroyed after each run). The agent cannot exfiltrate to arbitrary hosts.
What runs in the agent's environment?
- Your repo, checked out fresh
- The agent CLI (Claude Code or Codex)
- A shell + standard build tools (node, python, go, git)
- Per-workspace secrets you opt into
Nothing else.
Can I use my own LLM?
Roadmap: Aider with BYO API key (OpenAI-compatible endpoints). Today, only Claude or Codex.
Where can I follow updates?
- Changelog:
app.vemlor.com/changelog - Twitter/X:
@vemlorhq - Status:
status.vemlor.com