Shieldstral is a 3B open-weights multimodal safety classifier that outperforms larger models by up to 7x by framing content moderation as a policy-adaptive question-answering task. It accepts plain-language policies at inference time, unifying text and image safety evaluation without retraining, and is released under Apache 2.0, providing calibrated safety scores while operating efficiently on a single 16GB NVIDIA GPU.
mistral.ai
5 min
8/4/2026
Inkling is an efficient open model designed for text, images, and audio, featuring 975 billion total parameters with 41 billion active parameters using a Mixture of Experts architecture. It supports a context window of 64K to 256K tokens on Tinker and exhibits strong performance in knowledge, math, and science, along with capabilities for agentic coding and tool usage.
thinkingmachines.ai
1 min
7/15/2026
Muse Spark 1.1 is a multimodal reasoning model developed by Meta Superintelligence Labs, designed for agentic tasks with enhanced capabilities in tool usage, coding, and multimodal understanding. This release, alongside Muse Image, aims to improve performance efficiency and advance the vision of personal superintelligence.
ai.meta.com
6 min
7/9/2026
Gemini API's File Search tool now supports multimodal data and custom metadata for building retrieval-augmented generation (RAG) systems. The update includes page citations to enhance grounding and transparency, enabling better organization of text and visual content.
blog.google
2 min
5/10/2026
Gemma 4 delivers maximum compute and memory efficiency for mobile and IoT devices, enhancing intelligence in these platforms. It supports the development of autonomous agents capable of planning, navigating apps, and completing tasks, while also offering strong audio and visual understanding for rich multimodal applications and multilingual experiences that extend beyond simple translation.
deepmind.google
1 min
4/2/2026
Mistral Small 4 is the latest release in the Mistral Small family, unifying the capabilities of reasoning, multimodal processing, and agentic coding into a single model. This model eliminates the need for users to choose between different functionalities, providing a versatile solution for various tasks.
mistral.ai
5 min
3/16/2026
Shieldstral is a 3B open-weights multimodal safety classifier that outperforms larger models by up to 7x by framing content moderation as a policy-adaptive question-answering task. It accepts plain-language policies at inference time, unifying text and image safety evaluation without retraining, and is released under Apache 2.0, providing calibrated safety scores while operating efficiently on a single 16GB NVIDIA GPU.
mistral.ai
5 min
8/4/2026
Muse Spark 1.1 is a multimodal reasoning model developed by Meta Superintelligence Labs, designed for agentic tasks with enhanced capabilities in tool usage, coding, and multimodal understanding. This release, alongside Muse Image, aims to improve performance efficiency and advance the vision of personal superintelligence.
ai.meta.com
6 min
7/9/2026
Gemma 4 delivers maximum compute and memory efficiency for mobile and IoT devices, enhancing intelligence in these platforms. It supports the development of autonomous agents capable of planning, navigating apps, and completing tasks, while also offering strong audio and visual understanding for rich multimodal applications and multilingual experiences that extend beyond simple translation.
deepmind.google
1 min
4/2/2026
Inkling is an efficient open model designed for text, images, and audio, featuring 975 billion total parameters with 41 billion active parameters using a Mixture of Experts architecture. It supports a context window of 64K to 256K tokens on Tinker and exhibits strong performance in knowledge, math, and science, along with capabilities for agentic coding and tool usage.
thinkingmachines.ai
1 min
7/15/2026
Gemini API's File Search tool now supports multimodal data and custom metadata for building retrieval-augmented generation (RAG) systems. The update includes page citations to enhance grounding and transparency, enabling better organization of text and visual content.
blog.google
2 min
5/10/2026
Mistral Small 4 is the latest release in the Mistral Small family, unifying the capabilities of reasoning, multimodal processing, and agentic coding into a single model. This model eliminates the need for users to choose between different functionalities, providing a versatile solution for various tasks.
mistral.ai
5 min
3/16/2026
Shieldstral is a 3B open-weights multimodal safety classifier that outperforms larger models by up to 7x by framing content moderation as a policy-adaptive question-answering task. It accepts plain-language policies at inference time, unifying text and image safety evaluation without retraining, and is released under Apache 2.0, providing calibrated safety scores while operating efficiently on a single 16GB NVIDIA GPU.
mistral.ai
5 min
8/4/2026
Gemini API's File Search tool now supports multimodal data and custom metadata for building retrieval-augmented generation (RAG) systems. The update includes page citations to enhance grounding and transparency, enabling better organization of text and visual content.
blog.google
2 min
5/10/2026
Inkling is an efficient open model designed for text, images, and audio, featuring 975 billion total parameters with 41 billion active parameters using a Mixture of Experts architecture. It supports a context window of 64K to 256K tokens on Tinker and exhibits strong performance in knowledge, math, and science, along with capabilities for agentic coding and tool usage.
thinkingmachines.ai
1 min
7/15/2026
Gemma 4 delivers maximum compute and memory efficiency for mobile and IoT devices, enhancing intelligence in these platforms. It supports the development of autonomous agents capable of planning, navigating apps, and completing tasks, while also offering strong audio and visual understanding for rich multimodal applications and multilingual experiences that extend beyond simple translation.
deepmind.google
1 min
4/2/2026
Muse Spark 1.1 is a multimodal reasoning model developed by Meta Superintelligence Labs, designed for agentic tasks with enhanced capabilities in tool usage, coding, and multimodal understanding. This release, alongside Muse Image, aims to improve performance efficiency and advance the vision of personal superintelligence.
ai.meta.com
6 min
7/9/2026
Mistral Small 4 is the latest release in the Mistral Small family, unifying the capabilities of reasoning, multimodal processing, and agentic coding into a single model. This model eliminates the need for users to choose between different functionalities, providing a versatile solution for various tasks.
mistral.ai
5 min
3/16/2026
No more articles to load