Simple, honest pricing.
Every plan includes the full platform with usage included. No inflated token costs, no hidden multipliers. You see exactly what you pay for.
Free
No credit card required
- Full platform access
- Usage included to try everything
- Free models
- 1 vCPU / 1 GB per instance
- 1 concurrent agent
- Community support
Pro
For individual developers shipping with AI
- Everything in Free
- Generous usage included
- Frontier models (Claude, GPT, Gemini)
- 8 vCPU / 16 GB per instance
- Cloud agents & deployments
- 5 concurrent agents
- Ticket support
Pro+
For power users who need bigger instances
- Everything in Pro
- 2x usage included
- 16 vCPU / 32 GB per instance
- 10 concurrent agents
- Ticket support
Business
For teams that need governance and scale
- Everything in Pro
- 3x usage on all LLM models
- 16 vCPU / 64 GB per instance
- Unlimited concurrent agents
- SSO, RBAC & audit logs
- Priority ticket support
Enterprise
Custom pricing with volume discounts. For organizations that need dedicated infrastructure, advanced compliance, and white-glove support.
Radically transparent
50% of your subscription goes directly toward tokens and compute you actually use — at real provider prices, not inflated "credits."
We don't hide behind inflated token prices
Most AI tools mark up token costs 2–5x and call it “included usage.” You think you’re getting a deal, but you’re paying more per token than if you used the provider directly. We think that’s dishonest.
You pay the upstream provider rate for every token — the same prices listed on our AI Gateway models page. We don’t add markup.
See exactly how many tokens, compute minutes, and cloud resources you’ve used. Every dollar is accounted for, down to the request.
Usage is metered at the real provider-advertised cost. Not “1 credit = some opaque unit we defined.” What you see is what you pay.
Token prices reflect the real cost advertised by each provider, as listed on our models page. Standard payment processing fees apply to all transactions — these are set by payment providers (e.g. Stripe) and are outside our control. All payment methods carry an inherent processing cost.
Usage beyond your plan
When you exceed included usage, work continues at transparent per-unit rates. No hard cutoffs. Set spending limits to stay in control.
How organizations work
Each user in an organization has their own plan. Usage is tracked and billed individually — no shared limits, no surprise bills from a teammate.
Each member picks their own tier. A team can mix Free and Pro users — each gets the limits of their plan.
Every action counts against that user’s plan. One teammate’s heavy usage doesn’t eat into yours.
Enterprise customers can pool usage across all seats into one shared budget — better for large teams with uneven patterns.
Every plan includes the full platform
No add-on fees for features that other vendors sell as separate products.
Compare plans
| Free | Pro | Pro+ | Business | Enterprise | |
|---|---|---|---|---|---|
| Pricing | |||||
| Monthly price | $0 | $20 | $40 | $60/seat | Custom |
| Included usage | Trial | Generous | 2x Pro | 3x Pro | Pooled |
| Overage billing | — | Discounted | |||
| AI & Agents | |||||
| Models | Free only | Frontier | Frontier | Frontier | All |
| Concurrent agents | 1 | 5 | 10 | Unlimited | Unlimited |
| Shared AI chats | — | — | — | ||
| Compute & Cloud | |||||
| Max instance size | 1 / 1 GB | 8 / 16 GB | 16 / 32 GB | 16 / 64 GB | 32 / 128 GB |
| Cloud deployments | — | ||||
| Collaboration & Security | |||||
| Team members | 1 | 1 | 1 | Unlimited | Unlimited |
| SSO & RBAC | — | — | — | ||
| Audit logs | — | — | — | ||
| Pooled usage | — | — | — | — | |
| SLA | — | — | — | — | 99.9% |
| Support | |||||
| Support | Community | Ticket | Ticket | Priority ticket | Dedicated |
Frequently asked questions
How does included usage work?
Every paid plan includes a generous amount of usage for AI tokens, agent compute, and cloud services. Usage is metered at real provider costs with no markup. If you stay within your plan's included amount, you pay nothing extra. If you exceed it, overages are billed at the same transparent rates.
What happens when I exceed my included usage?
Your work doesn't stop. Overages are billed at the end of the billing cycle at the same per-unit rates. You can set spending limits to prevent unexpected charges — we notify you at 80% and can hard-stop at your cap.
What does "pooled usage" mean for Enterprise?
Instead of each seat having individual usage limits, Enterprise plans pool all usage into one shared budget. A team of 50 gets a combined allocation that any member can draw from. This avoids waste from underutilized individual quotas.
Can different team members have different plans?
Yes. Each user in an organization has their own plan. Some team members can be on Free while others are on Pro — each gets their own limits and is billed individually. One person's heavy usage doesn't affect anyone else.
Can I self-host instead?
Yes. Proxifai is open source under AGPL-3.0. You can deploy it on your own infrastructure at no cost. The plans above are for our managed cloud service — same codebase, vetted, hardened, and operated by us.
Stop paying for ten tools
One platform, one subscription. Start free, upgrade when you need more.