
Inkling 2 Mini
Thinking Machines Lab · US · #16 in our ranking · Best small model
Inkling 2 Mini is an open-weight mixture-of-experts model from Thinking Machines Lab (United States), released Nov 2026, with 26.4B parameters (3.4B active per token), a 1M-token context window, and the Apache 2.0.
Inkling 2 Mini is a looped mixture-of-experts model: the middle 12 of its 24 layers run three times, so it computes like a larger model while activating 3.4B parameters. The checkpoint is about 20GB and runs on one DGX Spark or a Mac with 32GB or more. It is released under Apache 2.0 with up to a 1M-token context.
Inkling 2 Mini specs
- Provider
- Thinking Machines Lab
- Origin
- US
- Architecture
- Mixture-of-experts
- Parameters
- 26.4B total · 3.4B active
- Context
- 1M tokens | 786K words
- License
- Apache 2.0 Commercial use
- Input
- Text, Image
- Released
- Nov 2026
- Weights
- Hugging Face
Sourced from the model card on Hugging Face. Last checked 2026-11-21.
Highlights
- 3.4B active parameters
- About a 20GB checkpoint
- Apache 2.0 license
Where it fits for payment teams
A fast, cheap model for high-volume steps that must stay on your own hardware, like tagging alerts or drafting case notes.
Harnesses in Runtime
Run Inkling 2 Mini through any of these harnesses on Runtime.
Provider-reported benchmarks
| SWE-bench Verified (bash only) | 79.2 |
|---|---|
| MCP Atlas (Public) | 82.7 |
| BrowseComp | 79.0 |
Benchmark scores are reported by each model provider on its Hugging Face model card. They are not independently verified by Runtime.
How to run Inkling 2 Mini
GPU requirements, Bedrock, and Vertex AIDownload the weights
Pull thinkingmachines/Inkling-2-Mini from Hugging Face. Check the Apache 2.0.
Serve it in your cloud
vLLM on your GPUs, or a managed host.
Put it to work in Runtime
Point an agent or skill at it, then compare with evals.
Inkling 2 Mini FAQ
What is the context window of Inkling 2 Mini?
+
Inkling 2 Mini supports a 1M-token context window (1,048,576 tokens), roughly 786K words of English text.
How many parameters does Inkling 2 Mini have?
+
Inkling 2 Mini is a mixture-of-experts model with 26.4B total parameters, of which 3.4B are active per token.
Is Inkling 2 Mini free for commercial use?
+
Yes. Inkling 2 Mini is released under the Apache 2.0 license, which is permissive and allows commercial use and fine-tuning.
What can Inkling 2 Mini take as input?
+
Inkling 2 Mini accepts text, image input and generates text.
Can I run Inkling 2 Mini in my own cloud with Runtime?
+
Yes. Download the weights from Hugging Face (thinkingmachines/Inkling-2-Mini) and serve them with vLLM in your cloud, or use a managed provider that hosts the model. Runtime agents can then use Inkling 2 Mini through a harness like OpenCode, with your data staying in your environment. Our how-to-run page for Inkling 2 Mini covers GPU memory, Bedrock, and Vertex AI.