Simple plans for every scale
Five plans, derived live from deploy/config/plans.yaml. Self-serve up to the Business tier, Dedicated above that.
Starter
$99/mo
Billed monthly
- 2 seats
- 250k input tokens / mo
- 100k output tokens / mo
- 1M embedding tokens / mo
- 1 GB document storage
- No custom domain
- SSO not included
- No SCIM
- No widget
- Models: small_pool
Pro
$499/mo
Billed monthly
- 10 seats
- 5M input tokens / mo
- 1.5M output tokens / mo
- 20M embedding tokens / mo
- 50 GB document storage
- 1 custom domain
- SSO not included
- No SCIM
- Embeddable widget
- Models: small_pool, medium_pool
Business
$1,499/mo
Billed monthly
- 25 seats
- 25M input tokens / mo
- 7.5M output tokens / mo
- 100M embedding tokens / mo
- 250 GB document storage
- 3 custom domains
- SSO not included
- No SCIM
- Embeddable widget
- Models: small_pool, medium_pool, large_pool
- Special-PI collections allowed
Dedicated (Managed)
$30,000/yr
Annual contract
- Unlimited seats
- Unlimited input tokens / mo
- Unlimited output tokens / mo
- Unlimited embedding tokens / mo
- Unlimited storage
- Unlimited custom domains
- SSO not included
- No SCIM
- Embeddable widget
- Models: *
- Special-PI collections allowed
Dedicated (On-Prem)
$60,000/yr
Annual contract
- Unlimited seats
- Unlimited input tokens / mo
- Unlimited output tokens / mo
- Unlimited embedding tokens / mo
- Unlimited storage
- Unlimited custom domains
- SSO not included
- No SCIM
- Embeddable widget
- Models: *
- Special-PI collections allowed
We meter input, output, and embedding tokens separately at model-tiered rates so your bill reflects what you actually use. Reranker tokens are folded into the base price.
| Plan | Input s/m/l $/M | Output s/m/l $/M | Embedding s/m/l $/M | Storage $/GB·mo |
|---|---|---|---|---|
| Starter | $0.50 / $2.00 / $8.00 | $1.50 / $6.00 / $24.00 | $0.10 / $0.20 / $0.40 | $0.20 |
| Pro | $0.40 / $1.60 / $6.40 | $1.20 / $4.80 / $19.20 | $0.08 / $0.16 / $0.32 | $0.15 |
| Business | $0.30 / $1.20 / $5.00 | $0.90 / $3.60 / $15.00 | $0.06 / $0.12 / $0.25 | $0.10 |
| Dedicated (Managed) | All-you-can-eat under annual contract | |||
| Dedicated (On-Prem) | All-you-can-eat under annual contract | |||
$10,000 for 60 days on a Dedicated Managed cluster. Fully credited toward year-one annual contract on conversion. Skip the free trial and prove the value with production-scale isolation.
Talk to sales →Can I bring my own model provider?
Yes — on an opt-in, per-collection basis. Bring your own key from any public cloud AI provider. We meter informationally; no overage charges on BYO traffic. The console surfaces a warning at opt-in time that the prompt leaves the feder8d boundary.
What happens when I hit a limit?
At 80% of any counter we send a warning. At 100% on a paid plan, overage kicks in. On trial, writes are blocked but reads still work.
Is there a free tier?
No free tier. We offer a 14-day trial on the Business tier — card optional. Hard usage caps apply during trial to prevent abuse.
Can I take the data with me if I leave?
Yes. One-click data export from the console produces a portable bundle (raw documents, vectors, chat history, config). On-prem customers run our restore CLI against it.