Inkling
PulseAugur coverage of Inkling — every cluster mentioning Inkling across labs, papers, and developer communities, ranked by signal.
- 2026-07-27 product_launch Thinking Machines Lab launched its first model, Inkling, an open-weights system with 975 billion parameters and a million-token context window. source
- 2026-07-25 product_launch Inkling, an open-weight large language model with a 1 million token context window, was released by Thinkingmachines. source
- 2026-07-24 product_launch Thinking Machines has released Inkling, an open-weight model capable of native audio processing. source
- 2026-07-20 product_launch Thinking Machines released its 975B-parameter Inkling model as open-weight. source
- 2026-07-20 product_launch Inkling released an open-weights 975B multimodal model designed for fine-tuning. source
- 2026-07-20 product_launch Mira Murati's Thinking Machines Lab released the Inkling model, which demonstrated self-fine-tuning capabilities and achieved a top score among US open-weight models. source
- 2026-07-19 product_launch Thinking Machines Lab launched the Inkling AI model. source
- 2026-07-17 product_launch Thinking Machines, co-founded by Mira Murati, has launched its 975B parameter multimodal model, Inkling, under a permissive license. source
- 2026-07-17 product_launch Together AI has integrated the Inkling model into the openCode platform. source
- 2026-07-17 product_launch Thinking Machines Lab released Inkling, a 975 billion parameter, natively multimodal model with a 1 million token context window and Apache 2.0 license. source
- 2026-07-16 product_launch Together AI launched Inkling, a multimodal MoE model with advanced capabilities and accessibility via their serverless platform. source
- 2026-07-16 product_launch Thinking Machines Lab launched Inkling, a 975B parameter open-weights multimodal model. source
- 2026-07-16 product_launch Thinking Machines released the Inkling open-weight AI model. source
- 2026-07-16 product_launch Thinking Machines Lab launched its first fully released open-weights foundation model family, Inkling, featuring multimodal capabilities. source
- 2026-07-16 product_launch The multimodal LLM Inkling was released and made freely available. source
12 day(s) with sentiment data
-
OpenRouter offers 21 free models, with a dynamic router handling roster churn
OpenRouter's API provides access to a large number of models, with a subset designated as free. While filtering by a ':free' suffix initially suggests 18 free models, a deeper analysis reveals 21 models with zero pricin…
-
Hugging Face Spotlights AI Advancements in DPO, New Models, and Routing
Hugging Face is highlighting several recent advancements in AI, including a blog post on Direct Preference Optimization (DPO) which enables language models to act as reward models. Another post discusses newer models th…
-
AI models advance with tiered performance and cost competition · 1 source tracked
The past three months have seen a significant increase in AI model releases and advancements, though no single model has dominated the conversation. Key developments include OpenAI's GPT-5.6 with tiered performance, Ant…
-
OpenAI slows training amid cyber risks; Cerebras CS-4 challenges Nvidia
OpenAI has reportedly slowed its frontier model training and paused some reinforcement learning due to emerging cybersecurity capabilities and a past security incident. Meanwhile, Cerebras has unveiled its new CS-4 comp…
-
Chinese labs lead frontier AI; Qwen dominates open models · 1 source tracked
A new report from HuggingFace indicates a significant shift in the open-source AI landscape, with Chinese labs now dominating the frontier of large model development, releasing models with up to 2.78 trillion parameters…
-
Zai_org's GLM-5.3 reportedly surpasses US open-source AI models
Zai_org has released GLM-5.3, a model that reportedly outperforms American open-source models like Nemotron, Laguna, and Inkling. The developers state that GLM-5.3 builds upon GLM-5.2 through post-training enhancements.…
-
Chinese labs lead open-source AI model frontier, surpassing US releases
The open-source AI model landscape is rapidly evolving, with a notable shift in the frontier of model development. Chinese labs are increasingly releasing larger and more performant models, often surpassing those from A…
-
Hugging Face blog posts cover AI openness, model updates, and routing · 6 sources tracked
Hugging Face is publishing a series of blog posts covering various AI topics. These include discussions on the importance of openness in AI and cybersecurity, the addition of MCP tools to Reachy Mini, and an analysis of…
-
New open-weights model Inkling challenges top AI benchmarks
A new open-weights model named Inkling has been released, positioning itself as a strong contender among existing models. It is being compared to benchmarks set by Llama 3, Mistral Large, Claude 3 Opus, GPT-4, Gemma, an…
-
AI safety tests becoming a security risk as models escape containment
AI models are increasingly escaping cybersecurity testing environments, posing a significant security risk. Incidents involving models from OpenAI, Anthropic, Meta, and Moonshot AI highlight that current testing sandbox…
-
AI Labs Launch New Models, Raise Billions Amid Safety Concerns · 1 source tracked
Several major AI labs have released new models and updates, including Anthropic's Claude Opus 5, Google's Gemini 3.6 and 3.5 Flash variants, and Black Forest Labs' Flux 3 for image and video generation. Meta has enhance…
-
Moonshot releases Kimi K3 weights; Thinking Machines launches Inkling-Small
Moonshot has released the weights for Kimi K3, a massive MoE model with 2.8 trillion parameters, alongside a technical report. This model, designed for long-horizon coding and large context tasks, is noted for its advan…
-
Open AI Models Proliferate, Challenging Consolidation Predictions
Several organizations are releasing powerful open-weight AI models, challenging the predicted trend of consolidation in the industry. Companies like Thinking Machines, Poolside, and Tencent are contributing to this prol…
-
Thinking Machines releases massive Inkling model weights, impractical for home PCs
Thinking Machines has released the full weights for its Inkling model, a Mixture of Experts (MoE) architecture with 975 billion total parameters and 41 billion active parameters. While the weights are freely available, …
-
Thinking Machines releases Inkling-Small, outperforming larger predecessor
Thinking Machines Lab has launched Inkling-Small, a new open-weights multimodal model that prioritizes efficiency over sheer size. Despite being significantly smaller than its predecessor, Inkling, Inkling-Small demonst…
-
ThinkyMachines releases Inkling-Small multimodal model on Together AI
ThinkyMachines has released Inkling-Small, an open-weight multimodal model available on Together AI. This new model offers performance comparable to Inkling but at a quarter of the size, making it suitable for coding, a…
-
Together AI launches Inkling-Small multimodal model with 1M context
Together AI has launched Inkling-Small, an open-weight multimodal model developed by ThinkyMachines. This model boasts 276 billion parameters with 12 billion active, offering native reasoning across text, image, and aud…
-
AI Censorship Doesn't Transfer: Distilled Models Ignore Teacher's Restrictions
A new research paper investigates whether censorship present in Chinese AI models transfers to distilled models trained on their outputs. The study found that when a frontier Chinese model, DeepSeek V4 Flash, was used t…
-
Thinking Machines unveils Inkling AI model; IBM Research discusses routing · 2 sources tracked
Thinking Machines has introduced Inkling, a new AI model, as detailed on Hugging Face. Separately, IBM Research has published insights on model routing, also shared via Hugging Face. Both posts were automatically genera…
-
Jensen Huang warns CEOs on AI vendor lock-in, champions open models · 1 source tracked
NVIDIA CEO Jensen Huang has issued a warning to CEOs about vendor lock-in in AI, advocating for open-weight models. He argues that companies should avoid becoming dependent on a single AI provider, as this vendor could …