Thinking Machines Launches Competitive Open-Weight AI Model
Inkling 2 Small challenges Chinese dominance in the open-weight artificial intelligence sector.
1New Western AI Contender
Thinking Machines Lab, founded by former OpenAI executive Mira Murati, has unveiled Inkling 2 Small. This model is currently proprietary but is scheduled for a public weight release next week. It represents a significant effort by Western labs to reclaim ground in the open-weight AI space, which has been dominated by Chinese developers.
2Performance and Technical Specs
The model features 594 billion parameters and supports image inputs while producing text outputs. It achieved a score of 31 on the Artificial Analysis Intelligence Index, placing it on par with Gemini 3.8 Flash. The developers specifically optimized the architecture to improve coding performance and integration within complex automated agent loops.
3Shifting Competitive Landscape
Chinese models like those from DeepSeek and Moonshot have recently outperformed expensive closed-source alternatives from major US labs. By releasing competitive open-weight models, Western companies aim to provide enterprise customers with viable alternatives to costly proprietary contracts. This shift forces investors to re-evaluate the long-term value of closed-source AI development strategies.
4Staggered Release Strategy
Thinking Machines is adopting a staggered release process, similar to the approach used by French lab Mistral for its Le Chonk model. This method involves extensive pre-release testing with cybersecurity and safety teams to mitigate potential risks. Such precautions have become standard practice for companies navigating heightened security concerns regarding AI deployment.
5Benchmarking and Accuracy Gains
Reviewers noted substantial improvements over the previous iteration, including a 41% increase in automation capabilities and a 20% boost in coding performance. Accuracy metrics also showed progress, with the hallucination rate dropping from 61% to 27% on the AA-Omniscience v1.1 benchmark. These gains highlight the rapid iterative development cycle currently defining the industry.
6Cost and Efficiency Trade-offs
Despite its technical prowess, the model faces criticism regarding operational efficiency. Artificial Analysis reports that Inkling 2 Small is notably slow and verbose during task execution. Furthermore, it is priced at $0.30 per million input tokens, which is higher than the $0.24 median cost for models within its specific performance class.
7The Open-Weight Civil War
The industry is currently experiencing a divide between open-weight and closed-source development philosophies. As Western labs increase their focus on open-weight releases, they are directly competing with established Chinese models that have set high benchmarks for accessibility. This competition is fundamentally altering how enterprises choose their AI infrastructure and software providers.
8Future Market Implications
The success of Inkling 2 Small will depend on how developers balance its high coding intelligence against its slower processing speeds and higher costs. As more Western labs adopt open-weight strategies, the market will likely see increased pressure on pricing and performance, ultimately benefiting users who require flexible, high-performance AI tools.
Original article: Latest news — Thinking Machines Launches Competitive Open-Weight AI Model