The same usage as Claude Code and ChatGPT — for less. Bolt runs on low-cost open models, so you get real agentic power without paying frontier-model rates.
Free
Free in beta
Try Bolt on our own models.
Add a teammate — team features turn on automatically.
There's no separate "Team" plan to buy. Invite a 2nd member to any paid plan and you get shared admin, pooled usage, and centralized billing — billed per seat, on one shared usage balance. Usage is metered and pooled, topped up with prepaid credits, and runs stop at zero — never a surprise invoice.
Enterprise
Custom
For organizations running Bolt at scale. Pooled consumption billed at cost + 5%, with committed-volume discounts and enterprise controls.
How you compare
Bolt is a discount to Claude Code and ChatGPT at equivalent token usage — not a price match.
| Plan | Bolt | Claude | You save |
|---|---|---|---|
| Pro | $12 / seat | $20 (Claude Pro) | 40% |
| Max | $60 / seat | $100 (Claude Max 5×) | 40% |
| Plan | Bolt | OpenAI | You save |
|---|---|---|---|
| Pro | $12 / seat | $20 (ChatGPT Plus) | 40% |
| Max | $60 / seat | $200 (ChatGPT Pro) | 70% |
Yes. Bolt is positioned as a discount to Claude Code and ChatGPT at equivalent token usage. Bolt runs on low-cost open Bang models, so the same agent workload costs a fraction of Claude's or OpenAI's published token rates. See the Claude and OpenAI comparisons.
Usage is metered and pooled across your team from one shared balance, topped up with prepaid credits. Runs stop when the balance reaches zero, so there is never a surprise invoice.
No. There is no separate Team plan to buy. Invite a second member to any paid plan and team features — shared admin, pooled usage, and centralized billing — turn on automatically, billed per seat.
Yes. Free is in beta and includes 1M input and 100K output tokens per month on Bolt's own models, then you can bring your own API keys.