Running Thinking Machines as a service, AI Infrastructure
Thinking Machines Agency Implementation, Cost-Efficient Multimodal AI Delivery
Learn how to build profitable retainer services around Inkling-Small's 276B-parameter model, leveraging variable reasoning effort and native multimodal processing to reduce inference costs by 20-40 percent while maintaining quality for client deliverables. Master fine-tuning workflows, agentic tool use, and pricing strategies that turn token efficiency into recurring revenue.
Open the decision record for Thinking MachinesWhat does running Thinking Machines for clients commit you to?
Published figures for this service. Blank fields are not published.
- Monthly tool cost
- Not published
- Time to first value
- Not published
- Payback
- Not modeled
- Guided implementation
- 8 hours
Is Thinking Machines worth running as a client service?
The evidence supports building a custom AI application service around Inkling-Small due to its open weights, multimodal reasoning, and efficient performance. However, the lack of published pricing and ROI data means agencies must validate costs and timelines internally.
An agency-fit judgement for reselling this service. It is separate from the tool description on the decision record.
Before you start
What has to be in place before the first client engagement.
Tools and subscriptions
- Tinker platform for fine-tuning
- Tinker Playground for chat testing
- Hugging Face for model weights and integration
- API/SDK access for custom integration
People and inputs
- ML engineering team to handle setup and fine-tuning
- Compute infrastructure capable of running 276B-parameter MoE model
- Access to NVIDIA GB300 NVL72 systems for training and inference
Estimated investment: Open-source model, no license fee; costs include compute, storage, and engineering time.
Included with the course
7 working documents for delivering this service.
- Reasoning Effort Pricing Matrix for Client Retainerstemplate
- Multimodal Workflow Setup Checklist (Audio, Image, Text)checklist
- Fine-Tuning Scoping Guide for Domain-Specific Modelsguide
- Token Cost Calculator: Inkling vs Inkling-Small ROIworksheet
- Agentic Tool Use Implementation SOPsop
- 1M Context Window Document Ingestion Workflowtemplate
- Self-Hosted Deployment and Hugging Face Integration Checklistchecklist
Listed by name. These documents are not yet published as individual downloads.