Inkling
Inkling has 975 billion total and 41 billion active parameters. It combines text, image and audio inputs and generates text.
What the release provides
The model uses sixty-six decoder layers and a mixture of routed and shared experts. Its model card describes up to one million tokens of context. Tinker’s current hosted limits depend on the selected service.Thinking Machines: Inkling model cardTinker: Current models and pricing
How to use it well
Practical guidance. Choose the input format and service limit before preparing a large document or recording. Multimodal understanding does not mean the model generates images or speech.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Downloadable, Provider product |
| Model identifiers | thinkingmachines/Inklingthinkingmachines/Inkling:peft:262144:sampling-nvfp4 |
| Context window | 262,144 tokens Tinker serverless inference beta; model-card maximum is larger Tinker: Current models and pricing |
| Input price | 1 USD per million tokens Tinker serverless beta, uncached input Tinker: Current models and pricing |
| Output price | 4.05 USD per million tokens Tinker serverless beta output Tinker: Current models and pricing |
| Weight license and packages | Apache 2.0; BF16 and NVFP4Thinking Machines: Inkling weight repository |
| Audio guidance | 16 kHz WAV; recordings under twenty minutes are recommendedThinking Machines: Inkling weight repository |
| API access | Serverless inference is beta; production use has a waitlistTinker: Current models and pricing |
Related models
- Inkling-Small
Thinking Machines Lab · Inkling
A smaller multimodal Inkling model with public weights.
Launch date not confirmedAPI · Downloadable · Provider product - Inkling 2 Small
Thinking Machines Lab · Inkling
A second-generation Inkling model available through Tinker.
Launch date not confirmedAPI · Provider product