Sign in

Pricing

Open source where you build.
Hosting when you need it.

Build, evaluate, and improve your agents, on your own infrastructure or with While.

Developer

For individuals and small teams.

FreeForever

  • Open-source SDK
  • Run simulations and evaluations
  • Behavioral analysis
  • Generate training data
  • Local training workflows
  • Community support
Get started

Hosting(optional)

Train, store, and run models without managing infrastructure.

Usage basedPay for what you use.

  • Model and artifact storage
  • Managed training (SFT / RL)
  • Hosted inference
  • Auto-scaling infrastructure
  • Global regions
  • Pay as you go
See hosting

Enterprise

For organizations with advanced needs.

CustomAnnual contract

  • Everything in Developer
  • Team collaboration
  • Version control and audit logs
  • SSO / RBAC
  • Private deployment (VPC / on-prem)
  • Dedicated support and SLA
Contact sales

Every rate is public. See the price book.

Compare plans

FeatureDeveloperFreeHostingUsage basedEnterpriseCustom
Open-source SDK
Simulations and evaluations
Behavioral analysis
Training data generation
Local training (SFT / RL)
Managed training (SFT / RL)add-on
Hosted inferenceadd-on
Team collaboration
Private deployment (VPC / on-prem)
SSO / RBAC
Dedicated support and SLA

Questions

Do I need a card to get started?

No. The SDK is free forever. Add a card under Account when you start hosting.

Can I run While. in my own environment?

Yes. The SDK runs on your machines, with your own keys and GPUs. Enterprise adds private deployment in your VPC or on-prem.

How does hosting pricing work?

Training is billed by the GPU hour at the rates in the open price book. An 8B model costs $0.24 a minute while it answers and $10 a month to keep. Importing a model is $10. A 4B model costs half.

Can I serve a LoRA adapter I trained?

Yes. Hand it to While. and we import it. Or import it into Bedrock yourself and register the ARN. Any OpenAI-compatible server you run registers by URL.

Does the served model score the same as the trained one?

Yes on the adapter we checked. It solved 35 of 100 held-out tasks through Bedrock. It solved 35 of 100 on vLLM. The difference is half a point and could be chance.

Can I use my own domain?

You get a subdomain on models.while.ai. Claim it once. For your own domain write to us.

What models are supported?

Training builds on open models such as Qwen. Evaluations run against any model, open or API.

Is there a minimum contract for Enterprise?

Enterprise is an annual contract sized to your usage. Talk to us and we will scope it in one call.

Hosting

Host your model.

You trained it. Put it behind an OpenAI-compatible endpoint on While. and get a subdomain of your own.

One URL for every model

Register once. Your key picks the account. The model name picks the row.

Runs where it runs

A Bedrock import in your AWS account or ours. Or any /v1 server you host. We store no key of yours.

Your subdomain

Claim acme.models.while.ai in one call. Only your keys work there.

Register

For a model in your own AWS account add a role named WhileModelsInvoke that trusts While. and can invoke the model. The endpoint assumes it on every call.

curl -X POST https://models.while.ai/models \
  -H "Authorization: Bearer $WHILEAI_API_KEY" -H "Content-Type: application/json" \
  -d '{"name": "nemotron-8b-t2s-r1",
       "arn": "arn:aws:bedrock:us-east-1:<account>:imported-model/<id>",
       "roleArn": "arn:aws:iam::<account>:role/WhileModelsInvoke"}'

Call it

Use the OpenAI SDK or curl or wai.Endpoint.

from openai import OpenAI

client = OpenAI(base_url="https://models.while.ai/v1", api_key="zp_...")
reply = client.chat.completions.create(
    model="nemotron-8b-t2s-r1",
    messages=[{"role": "user", "content": "Which customers never ordered?"}],
    stream=True,
)

Your subdomain

curl -X PUT https://models.while.ai/domain \
  -H "Authorization: Bearer $WHILEAI_API_KEY" -H "Content-Type: application/json" \
  -d '{"subdomain": "acme"}'
# https://acme.models.while.ai/v1 now answers for your account only

The served weights are the trained weights

We imported a text-to-SQL adapter into Bedrock. We scored it on the same 140 held-out tasks as the vLLM run.

Through the endpoint
35 of 100 tasks
On vLLM
35 of 100 tasks
The base model
26 of 100 tasks

The two servings are half a point apart. That could be chance. Both beat the base by 8 points. That is not chance. The recipe.

What we keep

  • Never a prompt. Never a completion. Never a request log.
  • Per model per day we count calls and errors and tokens and the five-minute windows it answered in. That is the bill.
  • Each account gets 600 requests a minute. Every call is over TLS behind a firewall.

What it costs

Three prices when While. hosts the model. A unit is a Bedrock Custom Model Unit. An 8B model is 2 units.

While it answers
$0.12
per unit-minute in five-minute windows
While it is kept
$5
per unit per month
Per import
$10
one time

Your own AWS account or server is free. An idle model wakes in about two minutes. The OpenAI SDKs wait for it.

Read the docs Ask about your own domain