Models·Thinking Machines

Inkling Small model review

thinkingmachines/inkling-small

An open-weight model that takes text, images, and audio and returns text.

$0.45Input, per million tokens
$1.20Output, per million tokens
524KContext window
262KLongest answer

524K tokens of context — about 393,216 words, on a scale that ends at 2M.

What it can do

  • Reasoning
  • Vision
  • Tool use
  • Open weights
  • Takes text
  • Takes image
  • Takes audio
  • Gives text

Released 30 Jul 2026 · checked 3 Oct 2026.

About it

Inkling Small is Thinking Machines’ open-weight model for text, image, and audio inputs, with text output. It supports reasoning and tools. Its context window is 524,288 tokens, and it can write up to 262,144 tokens in one answer. Usage costs $0.45 per million input tokens and $1.2 per million output tokens.

Review

In a line

Inkling Small is an open-weight model from Thinking Machines that accepts text, images, and audio, then responds with text.

What it's good at

It supports reasoning and tools, making it suited to tasks that call for either. It can also take images and audio as input, alongside text.

Price in plain terms

Input costs $0.45 per million tokens, and output costs $1.2 per million tokens.

Reading a 500-word page costs about 0.029 cents. A chat turn with about 1,000 tokens in and 500 out costs about 0.1 cents; 1,000 such turns cost about $1.05.

Context and limits

It can take in up to 524,288 tokens at once and produce up to 262,144 tokens in one answer. Its listed output is text, even though it can accept images and audio.

Ways to use it

The model supports tools and has open weights, which are model files available for others to use.

Verdict

Consider Inkling Small if you need a model that can take text, images, or audio and produce text, or if reasoning and tool support matter to your use. Its large context and output limits may suit work involving lengthy material. Skip it if you need image or audio output, since its listed output is text, or if you need confirmed details about specific apps or services that offer it.

Verdict

Inkling Small is worth considering for people comparing open-weight models that can take text, images, and audio, especially when reasoning or tool support is useful. Its large context and output limits may fit tasks involving extensive material. Look elsewhere if you need generated images or audio, or need to know which specific apps offer the model.

  • Open weights are available.
  • Accepts text, image, and audio input.
  • Supports reasoning and tools.
  • Can take up to 524,288 tokens at once.
  • Can produce up to 262,144 tokens in one answer.
  • Its listed output is text only.
  • No specific apps, chat services, or providers are named.
  • Input and output tokens have different prices.
  • Pick it ifPeople comparing open-weight models with image and audio input; Tasks that benefit from reasoning or tool support; Work involving long inputs or long text answers
  • Skip it ifYou need the model to generate images or audio.; You need confirmed details about specific apps or providers.

Pricing

Input costs $0.45 per million tokens and output costs $1.2 per million. The examples put a page at about 0.029 cents, a chat turn at about 0.1 cents, and 1,000 such turns at about $1.05.

Questions

Who makes Inkling Small?

Thinking Machines makes Inkling Small.

Can it understand images and audio?

Yes. It accepts text, image, and audio input, and its listed output is text.

How much does it cost?

Input costs $0.45 per million tokens; output costs $1.2 per million tokens.

How much text can it handle at once?

Its context window is 524,288 tokens. It can produce up to 262,144 tokens in one answer.

Does it support tools?

Yes, tool support is listed. Specific tools are not named.

Is it open-weight?

Yes. Its weights are available for others to use.

On these boards

More from Thinking Machines