Unofficial page — not affiliated with or endorsed by Thinking Machines Lab. Chat answers are AI-generated and may be wrong.
Thinking Machines Lab · Available now
Inkling 2 Small (2026): Thinking Machines Lab’s open-weight model, price and specs
Inkling 2 Small is Thinking Machines Lab’s second-generation model, announced on November 5, 2026: a 594B-parameter mixture-of-experts model with 17B active, a 1M-token context and image input, open weights under Apache 2.0, and $0.30 / $1.00 per million tokens on OpenRouter.
- Maker
- Thinking Machines Lab
- Type
- Text
- Released
- 2026-11-05
- OpenRouter id
- thinkingmachines/inkling-2-small
Per million tokens
Price
| - | Input | Output | Source |
|---|---|---|---|
| Via OpenRouter (thinkingmachines/inkling-2-small) | $0.30 | $1 | OpenRouter list price; Prime Intellect serves it at $0.15 / $0.50 · openrouter.ai |
Checked
Facts, with sources
- Thinking Machines Lab announced Inkling 2 Small on November 5, 2026, and made it available on its Tinker platform.Source: thinkingmachines.ai · checked 2026-11-14
- A mixture-of-experts model with 594B total parameters and 17B active, with image input and variable effort.Source: thinkingmachines.ai · checked 2026-11-14
- Text and image input, text output; context window up to 1M tokens; each token is routed to 8 of 384 experts.Source: huggingface.co · checked 2026-11-14
- The lab’s November 12, 2026 update says the weights are now available for download on Hugging Face; the model card lists the Apache 2.0 licence.Source: huggingface.co · checked 2026-11-14
- The lab reports a score of 31 on the Artificial Analysis Intelligence Index, above its earlier Inkling models, with gains in agentic coding, vision and knowledge work.Source: thinkingmachines.ai · checked 2026-11-14
- Inkling 2 Small is the only Inkling 2 model so far: the lab’s Hugging Face page lists it (NVFP4 and MXFP8 versions) next to the earlier Inkling and Inkling Small.Source: huggingface.co · checked 2026-11-14
- Listed on OpenRouter as thinkingmachines/inkling-2-small since November 9, 2026, at $0.30 / $1.00 per million tokens, with text and image input and a 1,048,576-token context.Source: openrouter.ai · checked 2026-11-14
- Six hosts serve it on OpenRouter: Prime Intellect at $0.15 / $0.50 per million tokens, and DeepInfra, Modal, Nebius, Together and Cloudflare at $0.30 / $1.00.Source: openrouter.ai · checked 2026-11-14
Use it now
Chat
FAQ
Questions
Is there an Inkling 2 besides Inkling 2 Small?
Not yet. Inkling 2 Small is the only Inkling 2 model Thinking Machines Lab has released; its Hugging Face page lists it next to the earlier Inkling and Inkling Small.
When was Inkling 2 Small released?
The lab announced it on November 5, 2026 on its Tinker platform. OpenRouter has listed it since November 9, and the weights went up on Hugging Face with the lab’s November 12 update.
How much does Inkling 2 Small cost?
On OpenRouter, $0.30 per million input and $1.00 per million output tokens; Prime Intellect serves it at $0.15 / $0.50. In the chat on this page a typical reply costs about 1 credit ($0.01).
How big is Inkling 2 Small?
594B total parameters with 17B active per token (8 of 384 experts), and a context window of up to 1M tokens.
Are the weights open?
Yes. Since the lab’s November 12, 2026 update the weights are on Hugging Face, and the model card lists the Apache 2.0 licence.
What are its benchmark results?
The lab reports 31 on the Artificial Analysis Intelligence Index, above its earlier Inkling models, with the biggest gains in agentic coding, vision and knowledge work.
Can it read images?
Yes. It takes text and image input and writes text.
Is this Thinking Machines Lab’s website?
No. This page is unofficial and not affiliated with or endorsed by Thinking Machines Lab. Every fact links to its source.