Deploy AI across your company.
Plan for the workloads
that come next.
One workspace for people, projects, keys and quotas. Plan models, capacity and procurement for distillation, data generation and batch workloads.
Turn your workload into a capacity plan.
Choose a model, rough scale and delivery preference. Exact numbers can come later.
Advanced mode uses separate exact inputs; it does not infer exact usage from a range.
Estimated requests:200,000 · Derived from scenario inputs; retries excluded
320M input + 160M output
- Required average throughput
- 111,111.1 TPM
- Required average request rate
- 46.3 RPM
- Estimated concurrency needed
- 24
These are workload requirements, excluding retries and peak headroom. Our team confirms deliverable capacity, model eligibility and your enterprise quote.
Discuss this workload planCoreBalanced cost & delivery
Claude Haiku 4.5 · 20251001
For generation and evaluation, start with Core to balance cost and consistency.
- Expected scale
- 500M–2B Tokens
- Target window
- Within 3 days
- Capacity status
- Confirmation required
Rough estimate at a 2:1 input/output mix and standard text rates, not a quote. Caching, request patterns and reserved capacity affect the final plan; completion windows need confirmation.
Discuss this planFreedom for teams. Control for administrators.
Understand usage ownership, then control each key’s quota, models and request sources.
Overview
Last 7 days
Usage trend
Sep 1–7 · Demo data; switch metrics and chart type
Project usage
Select a project to inspect keys and usage. Sample data only; no live account connection.
Hands-on help from pilot to rollout.
- You bring
- Share models, tools and team size.
- We help
- Map access, usage and capacity requirements.
Use-case brief & agreed scope
View handover checklist
- Target models and available tiers
- Users and applications
- Usage and launch goals
Business payments, invoices and procurement.
Confirm the receiving entity, bank details and payment reference before making a company transfer.
Discuss this payment method| Service description | Billing basis |
|---|---|
| AI API usage | Model × Tier × Usage |
| Enterprise volume pricing | As quoted |
Business invoices do not automatically include credit terms or jurisdiction-specific tax invoices. Confirm payment arrangements with our team.
What are you trying to do?
Choose your needs and a rough range, then leave an email. Company details and notes can wait.
A few things to confirm
What does distillation support include?
Capacity assessment, usage pricing and integration assistance for eligible model calls. Models and intended uses are subject to applicable terms and permissions. Data processing, training and student-model delivery require a separate assessment.
How is Enterprise priced?
Pricing starts with your model, available tier and usage, then accounts for capacity, completion windows and rollout. Different projects can use Flex, Core, Premium or Native without an additional fifth-tier plan.
Who qualifies for free onboarding?
Customers meeting an agreed enterprise usage level receive standard onboarding at no extra charge: workspace setup guidance, member and key allocation, permissions, quotas, supported client setup and admin handover. API usage is billed separately.
Can we use our company procurement process?
We support business invoices, company bank transfers, ACH and specified cryptocurrency payments. Our team confirms the receiving entity, currency and arrangements. Invoice availability does not automatically include credit terms or jurisdiction-specific tax invoices.
What can Workspace manage?
Organize access by project and key, review request usage, and set fixed or daily quotas, model access and IP allowlists. Member roles and project ownership help assign responsibility.
Do Native / Premium determine data handling?
Native describes first-party provider infrastructure. Data handling, retention, region and applicable terms must be reviewed for your use case; the tier alone does not determine these commitments.