All models / THINKINGMACHINES

thinkingmachines

Thinking Machines: Inkling Small (batch)

Name: Thinking Machines: Inkling Small (batch)
Creator: thinkingmachines
Published: 2026-07-30T20:25:17.000Z

Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of...

View source on OpenRouter Hugging Face

Overview

Model specifications

Context: 524,288 tokens
Maximum output: 471,859 tokens
Architecture: text+image+audio->text
Tokenizer: Other
Knowledge cutoff: Not provided
Moderated: No

Capabilities

Input → Output

Input

textimageaudio

Output

text

Reasoning

Yes · max · high · medium · low · minimal · none

API

Supported API parameters

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltystoptemperaturetool_choicetoolstop_ktop_logprobstop_p

Available providers

3 providers

DeepInfra

fp8

Context: 524K
Maximum output: 262K
Input: $0.450
Output: $1.20
Cache read: $0.100
Cache write: Not provided

BaseTen

fp8

Context: 1.0M
Maximum output: 33K
Input: $0.500
Output: $1.20
Cache read: $0.100
Cache write: Not provided

Together

unknown

Context: 524K
Maximum output: 472K
Input: $0.500
Output: $1.20
Cache read: $0.100
Cache write: Not provided

thinkingmachines

models

All models

thinkingmachines

Released Jul 31, 2026