More room in your API budget.
Draftworks brings a focused catalog of models into one API and passes lower infrastructure costs through to you.
Choose the model your work needs.
Use GPT-6 Astra, Claude Fable 5.1, Fable 5, Opus 5, Sonnet 5, Grok 4.6, Kimi K3, and GLM 5.3. Requests run through Microsoft Azure or Amazon Bedrock. Your model ID selects a specific model; Draftworks does not silently substitute another one.
Every model costs 40% below list. At $5,000 in lifetime purchases, Scale unlocks image and voice models at the same discount. Credits start at $5, and there is no subscription.
See what you paid for.
The dashboard records the selected model, cloud provider, token counts, and charge. Successful metered requests also get an Ed25519-signed receipt covering the model ID, token counts, and content hashes.
A receipt lets you verify our signature and detect changes to that record. It is a statement from Draftworks, not independent proof of the weights a cloud provider ran. Read the verification guide for the exact guarantees.
Know the limits before you start.
Endpoint and parameter support varies by model. Claude uses our Chat Completions adapter. Draftworks stores bounded request and response logs; Fable requests also require provider data sharing for safety review. The data-handling page explains retention.