// pricing

What each tier costs.

Every plan grants a credit balance you spend per token. Bring your own hardware and pay nothing per request. Storage stays while you are funded; only tiers with active infrastructure are shown.

Figures are published to signed-in accounts. Everything else is here: what each meter measures, how credits work, and what separates the tiers. Open an account to read the rates, or mail contact@erebine.ai and we will send them.

// service matrix

Every tier, every limit

A tier sets the price and limits. Hardware is set per endpoint, not per tier. Type to narrow columns by tier name or feature; press / to focus, Esc to clear.

Feature Free CPU AMD Optimized GPU NVIDIA Shared NVL Featured GPU NVIDIA Shared Featured GPU AMD Shared Self-Hosted
Pricing
Token billing Not billed Metered per token Metered per token Metered per token Metered per token Not billed
Hourly billing (EIM) - - - - - Billed by the hour
Request Rate Limit
Requests per minute 64 128 256 256 256 Unlimited
Capacity
Max model size 64 GB 20 GB Unlimited Unlimited Unlimited 512 GB
Endpoint Configuration
Concurrent requests 8 24 48 48 48 96
Max batch size 1 8 48 16 16 64
Request timeout 30s 600s 600s 300s 300s 1800s
Features
Streaming Yes Yes Yes Yes Yes Yes
Batching - Yes Yes Yes Yes Yes
Storage
Storage pool (platform-wide) Shared Shared Shared Shared Shared Shared
Minimum billable (display floor) 1 GB 1 GB 1 GB 1 GB 1 GB 1 GB
Content retention Retained as long as funded Retained as long as funded Retained as long as funded Retained as long as funded Retained as long as funded Retained as long as funded
Encryption at rest Yes Yes Yes Yes Yes Yes
// storage

Storage that bills the way you expect

One pool per project. One platform-wide rate. No per-content-type meters; no charges for bytes you do not store.

Per GB, per month

Storage is billed at a single platform-wide rate, not per service tier: base rate * markup multiplier, published in credits.

Display floor

The Usage dashboard reports billable storage as max(1 GB, actual usage). Per-cycle charges are prorated by the actual bytes you store, not the floor.

Shared per project

Models, completions, responses, conversations, batch files, and uploads share a single pool per project. No separate meter per content type.

Retained while funded

Content is retained as long as your project is funded. If credits reach zero, storage enters a 96-hour grace window; settle the balance within it and nothing is removed. A hot tier cache fronts recently used content for faster reads.

Encryption at rest

Stored content is encrypted with AES-256-GCM across both hot and cold tiers. Keys are managed per project for tenant isolation.

Metered from the first byte

No tier includes free storage. Usage is billed from the first byte you store, subject to the 1 GB display floor.

// subscription plans

Plans grant credits, not minutes

Each plan grants a credit balance every cycle and applies a per-token discount. The grant, not the price, is what you spend. The Free plan grants credits at no charge.

Plan Cycle grant Per-token discount
Free Credits, every cycle 1.00x
Starter Featured Credits, every cycle 0.95x
Hot-shot Credits, every cycle 0.85x
Extra Credits, every cycle 0.70x
Double Extra Deluxe Credits, every cycle 0.60x

One unit, every meter

Credits are the only unit. A plan grants them each cycle; tokens, storage, and hourly EIM spend them. A single published exchange rate converts credits to dollars, and every price on this page is a rendered consequence of that one rate.

Getting the numbers

Per-token rates, cycle grants, the storage rate, and the exchange rate are published to signed-in accounts. Open an account and they render here. Not signing up? Mail contact@erebine.ai and we will send them.

// sovereign

The whole stack, in your datacenter

The platform behind erebine.ai, licensed to run inside your network and on your accelerators. Full capability: nothing is removed, throttled, or held back for the private build. It is not a lesser edition of the hosted service; it is the hosted service, on your hardware.

per frontend node / month

Frontend node

The OpenAI-compatible API surface and the workspace UI your team works in.

per router node / month

Router node

The scoring router and the agent mesh: every dispatch decision and tool lease.

per agent / month

Agent

One license per inference or execution agent you run on your accelerators.

// included
  • Operational license for the full stack, granted per node.
  • Container images, packages, and binaries, ready to deploy.
  • Internal platform documentation and deployment instructions.
  • Direct access to engineering, the people who built the router and mesh.
  • 24x7x365 operations for inference and workspace.
  • Co-development on the integrations your deployment needs.
// at additional cost
  • Source code, licensed alongside the platform.
  • Deployment and system management, run hands-on by us.
  • Custom integrations, scoped per engagement.

Per node, billed monthly, six-month minimum. Operational licenses are granted per node. Term length beyond the minimum, isolation guarantees, and add-on scope are set per contract. Per-node rates are quoted on request: mail contact@erebine.ai.

Start the conversation
// questions

Read this before you deploy

Token math, tier semantics, rate-limit behavior, storage billing. Stated, not implied.

Token usage includes both input (prompt) tokens and output (completion) tokens. We use the same tokenization as the model you deploy, so counts match what you would see locally. You are only charged for successful requests.
Yes. On billable tiers, tokens served from the prefix cache are billed at 0.25x the standard per-token rate. The free and self-hosted tiers do not incur token charges, so the multiplier does not apply.
Yes. Erepress losslessly trims avoidable token bloat, insignificant whitespace and verbose machine-generated JSON, from every chat message before it is sent. The model reads the same text, so you are billed for fewer input tokens for the same result. It runs on every tier including the free one, at no extra cost, on by default, and adds no model call to your request. Turn it off per request, per workspace, or with a slash command.
The service_tier request parameter accepts flex, default, and priority. priority raises routing priority and is billed at 1.25x the standard token cost; flex and default carry no surcharge. Omit the parameter and you get default.
No, you cannot change an endpoint's tier after deployment. You can deploy new endpoints on different tiers at any time.
Requests exceeding your tier's rate limits receive a 429 (Too Many Requests) response with a Retry-After header. We recommend exponential backoff in your client. Consider a higher tier if you consistently hit limits.
No minimum commitment on any tier. Pay-as-you-go means you only pay for what you use. Stop or delete endpoints any time, no penalty.
All content types share a single storage pool per project. The Usage dashboard displays a 1 GB floor (max(1 GB, actual usage)), while per-cycle charges are prorated by the actual bytes stored. Storage is billed at a per-GB monthly rate administered platform-wide, not per service tier. Content is retained as long as the project is funded; at zero credits a 96-hour grace window precedes removal. Stored content is encrypted at rest with AES-256-GCM. See the storage documentation for full details.

Deploy a shared endpoint at zero per-request cost.

Move to dedicated hardware when the math says so. The free tier shares GPU; tier upgrades take seconds.

Open an account Read the docs