Pricing
Pay for work, not seats. No monthly commitments.
How credits work
Credits are consumed as your agents use LLM tokens. Different models have different per-token rates. You choose the model when dispatching an agent. Set a cost cap on any thread to prevent runaway spending.
Credits
- No monthly commitment
- Transparent per-model pricing
- Set per-thread cost caps
- No credit card required to start
- New users get $5 free
See current model pricing and usage details in the app after signing up.
Frequently asked questions
How do credits work?
Credits are consumed as your agents use LLM tokens (input and output). Different models have different rates. You can see the exact cost of each agent run in the app.
What happens when I run out?
Agent dispatches are blocked until you add more credits. Active sessions will complete their current inference call but won't start new ones.
Is there a free tier?
New users get $5 in free credits when they sign up — enough for several agent sessions. No credit card required.
Can I set spending limits?
Yes. Set a cost cap on any agent thread. The agent stops when the cap is reached, so you never get an unexpected bill.
How does this compare to Codex or Devin pricing?
Naiad charges per-token at transparent rates with no seat fees. You pay for actual model usage, not a monthly subscription. $5 goes a long way for typical tasks.