Mira Murati's Thinking Machines Lab Launches Inkling Model, Underperforms Chinese Rivals

Mira Murati, former CTO of OpenAI, has launched her own AI startup, Thinking Machines Lab, with a new model called Inkling. Inkling has 975 billion parameters and supports a 1 million token context window, but underperforms compared to leading Chinese open models like Zhipu's GLM 5.2 and Moonshot's Kimi K2.6.

Thinking Machines Lab released Inkling as an open model, available on their Tinker platform and Hugging Face, with open weights. The model is designed for customization and aims for well-rounded performance.

A group of publishers, including Hachette Book Group and Elsevier, have sued Google in a federal US court, alleging copyright infringement while training its Gemini AI models. The lawsuit claims Google used copyrighted books without permission.

Databricks is using AI to help higher education institutions scale limited resources and boost student success. The company's GenAI solutions integrate data ingestion, AI analysis, and insight discovery, all while maintaining strict data security and access controls.

AWS Marketplace has added new AI-based features to its product listings, anticipating continued growth in the use of enterprise agents. Enterprise agents are being used to research and make purchases.

Taiwanese companies listed on the TWSE and TPEx reported strong sales growth in June, driven by robust demand for AI applications. The semiconductor sector saw significant growth, with aggregate consolidated sales up 56.95 percent from a year earlier.

Key Takeaways

['Thinking Machines Lab launched Inkling, a 975B-parameter open model with a 1M token context window.', "Inkling underperforms compared to leading Chinese open models like Zhipu's GLM 5.2 and Moonshot's Kimi K2.6.", 'Publishers sue Google for copyright infringement while training Gemini AI models.', 'Databricks uses AI to help higher education institutions scale resources and boost student success.', 'AWS Marketplace adds AI-based features to product listings, driving growth in enterprise agents.', 'Taiwanese companies report strong sales growth in June, driven by AI applications.', "Inkling is available on Thinking Machines' Tinker platform and Hugging Face with open weights.", "Google's Gemini AI models allegedly used copyrighted books without permission.", 'Enterprise agents are being used to research and make purchases on AWS Marketplace.', 'The semiconductor sector saw significant growth, with sales up 56.95 percent from a year earlier.']

Murati's AI Startup Takes On China's Open Models

Mira Murati, former CTO of OpenAI, has launched her own AI startup, Thinking Machines Lab. Their new model, Inkling, underperforms compared to leading Chinese open models. Inkling has 975 billion parameters and supports a 1 million token context window. The model was pretrained on 45 trillion tokens spanning text, images, audio, and video. Despite its impressive specs, Inkling trails behind Chinese models like Zhipu's GLM 5.2 and Moonshot's Kimi K2.6 in benchmarks.

Thinking Machines Bets on Open Model Inkling

Thinking Machines Lab released its first open model, Inkling, which has 975 billion parameters and 41 billion active parameters. Inkling supports a 1 million token context window and was pretrained on 45 trillion tokens of text, images, audio, and video. The model is designed for customization and is available on Thinking Machines' Tinker platform. Inkling's performance is not the strongest among open models but aims for well-rounded performance.

Inkling: A 975B-Parameter Open-Weights Multimodal Model

Thinking Machines Lab released Inkling, a mixture-of-experts transformer with 975B total parameters and 41B active parameters. Inkling supports a context window of up to 1M tokens and was pretrained on 45 trillion tokens of text, images, audio, and video. The model is fine-tunable on Tinker and has open weights.

Inkling Tops US Open Models on Intelligence Index

Thinking Machines' Inkling model has become the top US open model on the Artificial Analysis Intelligence Index. Inkling has 975 billion total parameters and 41 billion active parameters. The model is efficient and performs well on various benchmarks, including GDPval-AA v2 and τ3-Banking.

Thinking Machines Releases Open-Weights Model Inkling

Thinking Machines Lab released Inkling, a general-purpose open-source model with 975 billion parameters. Inkling has a controllable effort setting, allowing developers to dial reasoning up or down. The model is available on Thinking Machines' Tinker platform and has open weights on Hugging Face.

Publishers Sue Google Over AI Copyright Infringement

A group of publishers, including Hachette Book Group and Elsevier, have sued Google in a federal US court, alleging copyright infringement while training its Gemini AI models. The lawsuit claims Google used copyrighted books without permission, despite internal warnings about the legal risks.

Google Sued Over Copyrighted Books Used for AI Training

A group of major publishers has sued Google, alleging the company illegally used millions of copyrighted books to train its Gemini artificial intelligence models. The lawsuit claims Google repurposed books originally supplied for limited services, such as Google Books and Google Scholar.

Small Businesses Can Grow With AI

Experts testified that small businesses using AI are likely to hire more employees, not fewer. AI can help small businesses focus on high-value work, automate routine tasks, and improve operations. The experts suggested that small businesses should focus on areas where AI can add value, start with solving specific problems, and invest in training workers on AI.

AWS Marketplace Drives AI Growth with Enterprise Agents

AWS Marketplace has added new AI-based features to its product listings, anticipating continued growth in the use of enterprise agents. Enterprise agents are being used to research and make purchases, and AWS is seeing significant growth in this area.

EU Forces Google to Share Search Data and AI Tech

The European Commission has announced new measures that will force Google to share its search data and AI technology with other companies. The measures aim to promote competition and innovation in the digital market.

Databricks Scales AI Support for Higher Education

Databricks is using AI to help higher education institutions scale limited resources and boost student success. The company's GenAI solutions integrate data ingestion, AI analysis, and insight discovery, all while maintaining strict data security and access controls.

Taiwan Listed Firms Report Strong Sales Growth

Taiwanese companies listed on the TWSE and TPEx reported strong sales growth in June, driven by robust demand for AI applications. The semiconductor sector saw significant growth, with aggregate consolidated sales up 56.95 percent from a year earlier.

Trader Claude's Portfolio Recovers

Trader Claude's portfolio recovered to $9,975 after exiting an Argentina World Cup Polymarket trade at +143.8pct and reloading NVDA at $208.25. The portfolio had a 13.3 pct single-day recovery.

Prediction Market Arbitrage Opportunities

Four live prediction market arbitrage opportunities were identified on July 16, 2026. These opportunities involve buying 'YES' and 'NO' contracts on different platforms, guaranteeing a profit regardless of the outcome.

Senator Josh Becker Hosts Virtual Town Hall on AI

Senator Josh Becker will host a virtual town hall on artificial intelligence and the future of work on July 16. The panel will discuss how AI is reshaping the workforce and what California can do to harness its benefits while protecting workers and consumers.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI Artificial Intelligence Machine Learning Deep Learning Natural Language Processing OpenAI Thinking Machines Lab Inkling AI Startup AI Model Multimodal Model Transformer Mixture-of-Experts AI Training AI Copyright Infringement Google Gemini AI AI in Business Small Business AI Adoption AWS Marketplace Enterprise Agents EU Competition Law Databricks AI in Education Taiwanese Companies AI Applications Semiconductor Sector Trader Claude Portfolio Recovery Prediction Market Arbitrage Senator Josh Becker Virtual Town Hall AI and Future of Work

Comments

Loading...