Meta released Muse Spark Open, an 816B total/42B active parameter open-weight model with a 1M-token context window, designed for coding and agentic work.
Apple’s MLX framework now supports local and distributed inference for large language models on Mac Studio with Apple silicon.
Anthropic’s Claude Code team describes a shift from code review as a final gate to continuous AI-assisted review.
Stack Overflow for Agents saw signups triple after integrating a ChatGPT plugin.
Elorian launched Glimpse, its first AI model, as a research preview with API and waitlist-only chat access.
Beam is a 501B-parameter sparse Mixture-of-Experts model with 23B active parameters per token and a 1M-token context window.
Meta released Muse Spark Open, an open-weight frontier multimodal mixture-of-experts model with approximately 816B total parameters and ~42B active per token.
Apertus 1.5 extends the 8B and 70B open-weight models with native image/audio input, a reasoning 'thinking mode,' and a 262,144-token context window.
Gemini 4 Argon is available globally for developers, enterprises, scientists, and consumers via API in AI Studio and the Gemini platform, Google AI Ultra subscribers in Antigravity, and AI Ultra/Pro subscribers in the Gemini app and AI Mode in Search.
Falcon ASR is a new 1.6B-parameter speech recognition model for Arabic, with emphasis on the Emirati dialect.
OpenAI suspended accounts tied to two influence operations that used its models to fabricate personas and organizations for geopolitical messaging.
Nemotron models fine-tuned to surpass gold thresholds in IOI 2026 and IMO 2026 competitions.
Playground lets users create, play, and share custom games using only text prompts, with no coding required.
OpenAI disclosed new results on open problems in mathematics derived from an internal frontier model.
A new analysis argues that current cyber-risk discourse ignores trade-offs between open and closed models, and misrepresents China’s risk posture.
Mistral AI released Mistral Large 4 (ML4), a 1 trillion-parameter multimodal model.
Falcon-Emirati-7B is a new 7B-parameter model specialized for Emirati Arabic, built on the Falcon-H1-Arabic base.
Z.ai’s GLM 5.3, a 753B-parameter mixture-of-experts model, is now available on Amazon Bedrock.
OpenAI details its planned approach to EU text provenance requirements using watermarking technology.
A local evaluation of the Qwen3.8 27B model tested its ability to add integers and return answers in English words across 5,070 cases.
pwasm 0.2a0 is a pure-Python WebAssembly engine released by Simon Willison.
Anthropic’s Frontier Red Team evaluated GLM-5.3 and Claude Mythos Preview on 100 randomly selected binary exploitation tasks.
Chatham Financial is using OpenAI's Codex and a newer GPT model to automate capital markets workflows.
Google AI detailed a slate of AI updates in September 2026, including the frontier model Gemini 4 Argon with a 1-million-token output limit and advanced reasoning capabilities.
OpenAI says it disrupted a coordinated campaign aimed at extracting protected model reasoning through adversarial distillation techniques.
Google DeepMind introduced Gemini 4 Argon, a frontier AI model designed for complex workflows in software engineering, enterprise knowledge work, and cybersecurity defense.
Gemini 4 Argon introduces an industry-leading 1 million token context window for sustained reasoning across complex, long-horizon workflows.
Gemini 3.8 Live with Live Avatar integrates real-time video streaming with speech for interactive enterprise agents.
OpenAI is granting the Government of Ukraine access to its Daybreak cybersecurity program to aid in defending civilian infrastructure.
Claude Opus 5.5 is positioned as the first in a new Claude 5.5 family, performing at the level of Claude Fable 5.1 for most tasks while costing 40% less to run than Opus 5.
Gemini 3.8 Flash TTS and Flash-Lite TTS are the latest additions to Google DeepMind’s text-to-speech lineup, designed for expressive audio generation and custom voice creation.
Jev is a new category of model called 'System One' or 'decision models' that outputs numeric decisions and confidence scores instead of text.
OpenAI announced an independent Advisory Group on Mathematics and Artificial Intelligence to guide the review and communication of emerging AI results.
Chinese open-weight models now lead U.S. models in downloads by about 1.6 billion, with 3.2 billion total downloads versus 1.6 billion for U.S. models.
Moonshot AI’s Kimi K3, a 2.8-trillion-parameter open-weight model, is now available on Amazon Bedrock.
Google disclosed that its Gemini model autonomously breached three real-world systems during internal security tests in May 2026.
Anthropic is consolidating its Claude Cowork and chat experiences into a single Claude interface.
OpenAI and AARP are partnering to offer free, in-person ChatGPT workshops for older adults across 10 U.S. cities.
Google outlined plans to support AI in 1,000 languages, up from 300 today, emphasizing direct audio understanding over text translation.
Google launched AI & Economy ATLAS v1.0, an interactive open-access tool visualizing global AI adoption patterns across occupations and countries.
Gemini 3.8 Live and 3.8 Live Extended Thinking are designed for fluid, real-time voice interactions with advanced reasoning and visual grounding.
OpenAI announced GPT-6 Astra, a next-generation model it says is state-of-the-art at navigating computers, web browsers, coding, and math.
OpenAI published a randomized study involving more than 1,000 students to assess the impact of ChatGPT on learning outcomes.
ChatGPT can now connect to trusted healthcare data sources, including EHR systems.
A curated reading list compiles research, essays, and policy analyses on open models published between 2023 and 2026.
Anthropic released Claude Fable 5.1 and Mythos 5.1 as its new flagship models for coding and knowledge work, positioning them as the world’s most advanced in these domains.
Anthropic released Claude Fable 5.1 and Mythos 5.1, addressing customer feedback on pricing, data retention, and safeguards.
Gemini 3.8 Flash improves reasoning and coding performance over 3.7 Flash while maintaining the same speed and cost.
Google released Gemini 3.8 Flash, positioning it as a harder‑working successor to Gemini 3.7 Flash by performing more reasoning steps and calling tools iteratively on complex tasks.
Qwen released Qwen3.8-Flash-Next as an open-weights multimodal mixture-of-experts model.
Tencent released Hy4 Preview, an open-weight text-only LLM with 770B total parameters and 49B active parameters.
OpenAI’s ChatGPT Work Cloud is accessible only to paid subscribers ($20/month and up) and offers features not available in standard Chat.
The ChatGPT desktop app includes a 1.7 GB local runtime with Python, Node.js, and native binaries for LibreOffice, Poppler, and git.
Anthropic published updated system prompts for its consumer-facing Claude models, including Fable 5.1, Haiku 4.5, and others.
NeoMME is a family of multilingual multimodal encoders with 260M and 800M parameters, trained from scratch without a separate vision tower or language model.
OpenAI announced it has scaled its online storage infrastructure to serve over 1 billion ChatGPT users.
IBM released Granite Time Series PatchTST-FM-r2, a ~385M-parameter time-series foundation model under Apache 2.0 and OpenMDW 1.0 licenses.
Gemini Omni 1.1 Flash introduces production-ready generative video controls for developers via the Gemini API in Google AI Studio.
Z.ai confirmed it built Ox Alpha, a new open-weight reasoning model that currently leads public benchmarks.
Granite 4.2 is IBM’s first family of dense, decoder-only reasoning LLMs, released in three sizes: 3B, 8B, and 30B.
OpenAI is expanding its ChatGPT for Teachers program to 55 U.S. school districts.
OpenAI’s CFO outlined a strategy linking advances in chips, compute, models, and products to scale AI capabilities.
OpenAI terminated accounts linked to a Russia-origin influence campaign that used AI-generated content.
A SQLite database file can be formatted to contain an ELF executable by storing its components in SQLite tables.
Linus Torvalds described an AI tool as a "tireless helper" during a challenging Linux kernel debugging session.
Promptwatch reports that site: operator usage in ChatGPT Search queries jumped from 0.3–0.5% to 16–17% on Aug 8, aligning with the GPT-5.6 rollout.
OpenAI reaffirms its zero data retention policy for eligible API customers using frontier models.
OpenAI reaffirmed its zero data retention policy for eligible API customers.
Nvidia is reportedly spending $26 billion to foster an open-model ecosystem where companies build their own models rather than rely solely on closed providers like Anthropic or OpenAI.
Jais 2 is a family of Arabic-centric open LLMs released by MBZUAI, Cerebras, and Inception, including a 70B-parameter model and an 8B-parameter variant.
Qwen 3.8 27B is an Apache 2.0–licensed 27B-parameter vision-capable LLM from Alibaba’s Qwen lab.
Z.ai’s GLM-5.3 matches or exceeds leading models like Kimi K3, Claude Fable 5, and GPT-5.6-Sol on agentic coding benchmarks despite having ~750B parameters—about a third the size of Kimi K3.
Gemini 3.7 Flash is positioned as Google DeepMind’s most capable workhorse model for coding and agentic workflows.
DeepSeek’s V4 Pro 0813 model is accessible via API on OpenRouter, with no official announcement from DeepSeek.
LiquidAI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model designed for on-device and edge applications.
OpenAI and AWS are making Daybreak cybersecurity models available via Amazon Bedrock for enterprise use.
AMIE is a research medical AI system developed by Google Research and Google DeepMind.
Meta released Muse Glimmer, a 30-billion-parameter open-weight vision model under an Apache 2.0 license.
A wave of startups is challenging transformer dominance in LLMs by proposing alternatives to dense attention mechanisms.
Auto mode will be the default setting for new Claude Code sessions in Pro, Max, and Team plans beginning August 14, 2026.
Accenture’s agentic AI strategy lead cited internal data showing non-engineers drive high token consumption.
Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.
Claude Fable 5 completed a 3D browser game from a four-year-old prompt and two reference images without further human input.
Adds support for claude-fable-5, claude-sonnet-5, and claude-opus-5 through LLM 0.32.
Google announced three new Gemini models—Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber—designed for scaling agentic workflows.
MiniMax-H3 is an omni-modal model that accepts text, images, audio, and video to generate up to 15-second video clips with audio.
OpenAI announced GPT-Live, a realtime voice AI system designed for continuous, turnless conversation.
Alibaba released Qwen3.8-Max, its largest and most capable AI model to date, claiming performance that rivals Anthropic’s Claude Fable 5 and OpenAI’s top systems.
Interconnects launched the Artifacts Hub, a curated index of 792 open models released in the last two years, with intelligence and adoption metrics.
Poolside released Laguna S2.1, an updated 118B-A8B MoE model optimized for DGX Spark systems.
DeepSeek’s V4-Flash-0731 is a 304-billion-parameter model with substantially enhanced agentic capabilities.
OpenAI disrupted a Cambodia-based scam operation that allegedly used ChatGPT to support multiple fraudulent schemes.
OpenAI is granting 100,000 academic researchers free access to its most advanced ChatGPT models.
Lyria 3.5 is a new music generation model from Google DeepMind now available in Google Flow Music.
Claude Mythos Preview was used for 60 hours (~$100,000 in estimated API costs) to probe HAWK and a weaker AES variant for cryptographic weaknesses.
OpenAI reports early deployments of AI coding agents in scientific computing workflows.
Moonshot released Kimi K3, a 2.8 trillion parameter model available as open weights on Hugging Face.
Google DeepMind introduced three new Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, designed to improve efficiency, latency, and reliability for AI agents.
A 28.9M-parameter language model runs entirely on an $8 ESP32-S3 microcontroller, generating text at about 9.5 tokens/sec without server support.
Claude Opus 5 is available today as a cost-efficient upgrade to Anthropic's Opus family, delivering near-frontier intelligence at half the price of Claude Fable 5.
Anthropic released Claude Opus 5, positioning it as a leader on the Artificial Analysis Intelligence Index’s agentic knowledge work benchmark.
Anthropic released Claude Opus 5, described as a thoughtful and proactive model that approaches the frontier intelligence of Claude Fable 5 at half the price.
OpenAI is expanding ChatGPT with Health features for eligible U.S. users.
Poolside AI, a small research team, claims its Laguna S 2.1 model (118B total parameters, 8B activated per token) outperforms larger open-weight models, including a ~1T parameter model from Thinking Machines.
Gemini Intelligence now automates tasks across more than 40 popular apps on new Samsung Galaxy Z Fold8 and Flip8 devices.
Gemini 3.6 Flash reduces output token usage by 17% compared to 3.5 Flash and improves coding, knowledge work, and multimodal performance.
Google DeepMind released three new Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, emphasizing efficiency, latency, and reliability for AI agents.
Claude Tag, Anthropic’s collaborative Slack-integrated coding agent, now lands 65% of product engineering PRs for the Claude Code team.
OpenAI describes new safety risks tied to long-horizon AI models in a newly published report.
Moonshot AI released Kimi K3, a 2.8 trillion parameter Mixture-of-Experts model slated for open-weights release on July 27, 2026.