Congrats to @thinkymachines on the release of Inkling-small! A smaller variant of Inkling, now live on our AudioMultiChallenge and MCP Atlas leaderboards. Inkling-small is tied for🥇on AudioMultiChallenge, scoring about the same as the larger Inkling despite the size difference. Strong multi-turn audio performance has typically come from the biggest models, so this is a promising signal for teams building voice applications where latency and cost matter. Also notable, on MCP Atlas it ranks second among open models on tool calling, behind Kimi K3 and ahead of GLM 5.2. Holding up on both audio reasoning and tool use at this size is a strong showing.

Whereas I felt like it took a village to release inkling, inkling-small felt much more routine 😆 We just took the pipeline used for Inkling, passed in a smaller model, and voila - new model! Inkling small benefited quite a bit vs Inkling from some minor improvements, but there's still so much more left in the tank...

The knowledge that makes AI useful is diffused. It lives with scientists, engineers, clinicians, firms. For AI to benefit from distributed knowledge, it must itself be distributed. Agree with Jensen that this is a future worth building.

Congrats to @thinkymachines on the release of their open weight model Inkling! We were proud to work with their incredible team on preparing this model for release for the past several months. Now live on our MCP Atlas and AudioMultiChallenge leaderboards. Inkling tied for 🥇 on AudioMultiChallenge, surpassing Gemini 3 Pro as the de facto frontier model that supports native audio input. Also notable, on MCP Atlas Inkling had a low hallucination rate compared to other frontier models.

It truly takes a village to release a model, perhaps especially an open weights model. Actually doing the entire process from scratch, from data to pretraining to posttraining to actual release, gives a lot of appreciation for anyone who does it! There's so many places to go wrong, and indeed so many things we would (and will!) do differently for a new model. But I'm happy with where we ended up :) Onto what's next!

Thinking Machines has released Inkling, the new leading U.S. open weights model, debuting at 41 on the Artificial Analysis Intelligence Index @thinkymachines has previously released research previews of models and this is their first production language model release. The model is 975B total parameters, has 41B active parameters, and accepts text, image, and audio input modalities. The model is accessible via Thinking Machines’ Tinker platform API (256K context window) and weights are available on HuggingFace (1M context window). Key results: ➤ Inkling debuts at 41 on the Artificial Analysis Intelligence Index, making it the leading open weights release from a U.S. lab. Inkling scores 3 points higher on the Intelligence Index (41) than the previous leading U.S. open weights model, Nemotron 3 Ultra (38), and also beats Gemma 4 31B (29) and gpt-oss-120b (24) ➤ Inkling stands out on agentic performance. It scores higher than both Kimi K2.6 and DeepSeek v4 Flash on both GDPval-AA v2 and 𝜏³-Banking: Inkling scores an Elo of 1238 on GDPval-AA v2, higher than Kimi K2.6 (1190) and DeepSeek v4 Flash max (1189) and scores 24% on 𝜏³-Banking, higher than Kimi K2.6 (21%) and just above DeepSeek v4 Flash max (23%) ➤ Inkling is token efficient compared to open weights leaders. Inkling averages 25K output tokens per Intelligence Index task compared to 43K, 38K and 37K by GLM-5.2 (max), Kimi K2.6 and DeepSeek v4 Pro (max) respectively ➤ Inkling natively supports image and audio multimodal inputs, a key differentiator among open weights models. Inkling accepts text, image, and audio input modalities. Images and videos are encoded via a hierarchical patch encoder and audio via discrete token encoding, with all modalities projected into a shared hidden space and processed jointly by the decoder Additional model details: ➤ Size: 975B (41B active) parameters ➤ Input modalities: Text, image, and audio (text output) ➤ Context window: 256K tokens on Tinker, open weights model supports 1M ➤ Pricing per 1M tokens (64K context window): $1.87 input / $0.374 cached / $4.68 output ➤ Pricing per 1M tokens (256K context window): $3.74 input / $0.748 cached / $9.36 output

Today we share the worldview behind our mission. Human values don't average out. Local knowledge can't be centralized. The good future has many AIs, raised in different places, shaped by the people they serve, disagreeing with each other the way we do.

What do we do at @thinkymachines: Personalization/sovereignty, Human Participation, Decentralization. Democratize AI and make it useful for people. All three of them reduce society's dependence on centralized AGI companies (including ours when we get important), and that is a future worth aiming for. You've seen a preview of this with Tinker, Interaction models and our research openly published on Connectionism. A **lot** more to come very very soon...

Thinking Machines exists to pursue a differentiated view of the future of AI. Specifically, we care about those branches of the tech tree that are involved with maintaining the best of human participation in the new world in which we find ourselves. It is our belief that this is the most important thing we can do with our time and talents. To us, this mission is both moral and opportunistic. We believe this is a moral imperative, because we think humans are important, and the best future is one where humans and AI work together deeply and functionally, not where humans are alienated from our work and lives by mass replacement. We also think this is practical and economic: AI that is built for active human participation produces better work, builds a more resilient world, and is a more compelling version of AI safety than systems optimized to operate apart from us. The narrative of mass automation and disempowerment that dominates the frontier of AI today can feel like a black hole. We, tiny astronauts, are dragged towards it—a fearsome maw amidst an expanse of glimmering potential. In that future, machine intelligence is built to act on its own and to replace us; we are relegated to observers and consumers of a progress we can no longer touch. A small circle owns the systems, and everyone else is left behind, watching from afar. That road may produce enormous, concentrated material abundance; it will certainly create a world in which most people have no real part. At Thinking Machines, we remember that the universe of possible futures is large, and understand we have not yet crossed an event horizon. There is yet another way.

We started Thinking Machines a year and a half ago with a couple of instincts: that people should have much more ability to customize models and do research on them, and that even as AI becomes more autonomous, there's a lot more to build to make humans and AIs work well together. A lot has happened since then, especially the massive progress in agents, so we wanted to revisit those instincts in light of everything we've learned, argue about them, and write down what we actually believe now. This is where we landed after a lot of debate. I'm happy with it!

1. (System design) - The Interaction Models see your screen and collaborates with you live. Here we're building a scalable system architecture together — no copy-pasting, no switching tabs, just thinking out loud and drawing on the screen together.