📊 Full opportunity report: The First Signs Of AI Change: Insights From Thinking Machines’ Inkling on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Thinking Machines announced the release of Inkling, a 975-billion-parameter open-weight AI model, emphasizing transparency and openness. The model is not the strongest but demonstrates a new approach to AI deployment and licensing, raising questions about open source and usage policies.
Thinking Machines has officially released Inkling, a 975-billion-parameter, open-weight multimodal AI model, available on Hugging Face under the Apache 2.0 license. This move signals a shift towards greater transparency and openness in the AI industry, contrasting with many proprietary models.
Inkling is a Mixture-of-Experts transformer with 975 billion total parameters and 41 billion active, supporting a 1-million-token context window. It was trained on 45 trillion tokens encompassing text, images, audio, and video, with a native multimodal input design that processes text, images, and audio jointly without an external vision adapter.
Unlike many recent models, Inkling’s weights are openly available on Hugging Face under Apache 2.0, allowing anyone to download, modify, and deploy independently. The model was trained using a hybrid optimizer on NVIDIA GB300 systems, with over 30 million reinforcement learning rollouts improving its reasoning capabilities.
However, the model’s release is accompanied by a Model Acceptable Use Policy (AUP), which reportedly restricts surveillance, deception, and automated decision-making affecting individuals, raising questions about the true openness of the model’s licensing and use restrictions.
The weights came first: what Inkling actually signals
Mira Murati’s lab shipped its first foundation model — and the model isn’t the story. The order of operations is: full weights, Apache 2.0, day one, before any closed API. Plus a rare concession — the lab says it’s not the strongest model available, open or closed.
- AIME 2026 97.1%
- GPQA Diamond 87.2%
- MCP Atlas (Nemotron 44.7%) 74.1%
- VoiceBench · open-weight audio frontier 91.4%
- FORTRESS adversarial · best open 78.0%
- ForecastBench · calibration 61.1
- HLE text-only (GLM-5.2 40.1%) 29.7%
- SWE-bench Pro (GLM-5.2 62.1%) 54.3%
- Terminal-Bench 2.1 (GLM-5.2 82.7%) 63.8%
- SWE-bench Verified (Fable 5 95.0%) 77.6%
- Design Arena · 2nd open, behind GLM-5.2 ~10th
A 0.2 → 0.99 effort setting trades reasoning tokens against cost & latency, so you get a curve, not a point. On Terminal-Bench 2.1 it reportedly matches Nemotron 3 Ultra at ~⅓ the tokens. Peak score is a vanity metric when you serve millions of calls; the cost curve is what ships. (Bonus: its chain of thought compressed on its own during RL — nobody rewarded it; efficiency did.)
Pitched as the Western alternative to Chinese open weights (censorship-resistance training is the differentiator). But GLM-5.2 still wins on agentic/reasoning and Kimi K2.6 often on multimodal: best American open model, second in the open field. The irony — post-training was bootstrapped on synthetic data from Kimi K2.5.
BF16 needs ≥2 TB aggregate VRAM (8× B300 / 16× H200). NVFP4 still needs ≥600 GB. Not a workstation model — a 512 GB fleet falls just short. “Open” ≠ “runnable.” Mitigations: 1-bit GGUFs (~74% acc.), hosted eval routes, and Inkling-Small (12B active) — the release local-first builders actually want.
Open weights used to be a consolation prize. Inkling is a strategic open release — Apache 2.0, natively multimodal, honestly marketed, published complete on day one, optimized for deployment rather than headlines (the model isn’t the product; the fine-tuning platform is). It doesn’t need to win every benchmark for that to matter. The frontier is learning that owning the base beats renting the API — arriving now from the inside. For the sovereignty buyer: ① a real Western hedge against being switched off · ② verify the use policy before you build · ③ check the VRAM, then benchmark vs GLM-5.2 & Kimi K2.6 on your task.
Implications of Open-Weight Release in AI Industry
The release of Inkling under open weights represents a notable departure from the trend of proprietary models, potentially enabling broader access and innovation. It underscores a shift towards transparency, but the accompanying use restrictions highlight ongoing tensions between openness and control, impacting how organizations may adopt and trust such models. This move could influence future open-source AI development and licensing practices, emphasizing the importance of clarity around licensing and usage policies for responsible AI deployment.
Domain-Specific Small Language Models: Efficient AI for local deployment
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background of AI Model Releases and Industry Trends
Over recent years, many leading AI models have been released as closed-source, with proprietary weights and restricted licenses, limiting independent scrutiny and customization. The industry has seen debates over transparency, safety, and control, especially following incidents where models were shut down or restricted due to regulatory or ethical concerns.
Thinking Machines, founded by former OpenAI CTO, has taken a different approach by releasing Inkling’s weights openly, albeit with some restrictions. This aligns with a broader movement advocating for open models, but also raises questions about the balance between openness and responsible use.
“Open weights are not the same as open source. The restrictions layered on top could limit real openness.”
— Industry observer, on X
HIWONDER Humanoid Robot with ChatGPT Multimodal AI Models AI Embodied Intelligent Vision Scene Voice Understanding 20DOF Educational Robotic Kit Python Programming, TonyPi Pro & RaspberryPi 5 16GB
Al-Driven & Raspberry Pi Powered. TonyPi is a high-performance AI vision robot designed for AI education applications. It…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Questions About Inkling’s Use Restrictions
While the weights are openly available, reports suggest that Thinking Machines enforces a separate Model Acceptable Use Policy (AUP) that restricts surveillance, deception, and certain decision-making applications. The exact scope, enforceability, and transparency of this policy remain unverified, raising questions about the true openness of the model.
It is not yet clear how these restrictions will impact commercial or research use, or whether they will be enforced uniformly across different jurisdictions.
Ollama – Open Source AI Tool Complete Guidebook
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for Adoption and Policy Clarification
Expect independent researchers and organizations to scrutinize Inkling’s licensing and restrictions closely. Further testing and benchmarking will likely follow to assess its safety, robustness, and compliance with stated policies. Additionally, more details about the Model Acceptable Use Policy are anticipated to clarify the boundaries of permissible use.
Thinking Machines may release updated versions or clarify the licensing framework, influencing how open-source models evolve in the industry.
AI WEALTH WORKFLOW 2026: A step by step guide to building real wealth with AI in 2026
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What makes Inkling different from other large language models?
Inkling is notable for its open weights under Apache 2.0 license, multimodal input support, and a focus on transparency, although it reportedly includes usage restrictions via a separate policy.
Is Inkling truly open source?
The weights are openly available under Apache 2.0, but reports suggest that a separate Acceptable Use Policy imposes restrictions, which complicates the notion of full openness.
Why does the licensing matter for users?
Licensing determines what users can do with the model—whether they can modify, deploy, or commercialize it freely. Restrictions could limit applications, especially in sensitive domains like surveillance or automated decision-making.
What are the potential risks of using Inkling?
Risks include uncertainties around enforcement of use restrictions, possible limitations on deployment, and ethical concerns if restrictions are not transparent or are difficult to comply with.
What impact could Inkling have on the AI industry?
Its open weights could encourage more open development and experimentation, but the layered restrictions highlight ongoing debates over balancing transparency with control and safety in AI deployment.
Source: ThorstenMeyerAI.com
