TBC puts a neuron-derived video model into public APIs
The Biological Computing Co. released mischief 1 for short image-to-video generation on October 30.
AI news by topic
AI model releases and product news
Model launches can change how people write, code, search and build software. This section follows new releases and product updates, with attention to what is available, who can use it and which capabilities are still being tested. Read the underlying announcements and evaluation sources alongside each briefing.
Latest reporting
43 briefings
Nov 1, 2026
DeepSeek releases a proactive coding-agent research prototype
A new DeepSeek GitHub repository explores a coding agent that monitors work and initiates tasks.
Oct 31, 2026
Google unveils Gemini 4 Carbon with a restricted cyber-defense rollout
Google's next frontier model starts with selected cyber defenders rather than general access.
OpenAI brings Astra weights to GPT-Rosalind-6 under trusted research access
OpenAI's October 30 system card describes a new Rosalind deployment using unchanged GPT-6 Astra weights.
Oct 29, 2026
Mistral releases Large 4 weights after a three-week public preview
Mistral has released the improved weights for its 1-trillion-parameter Large 4 model, adding self-hosted deployment, three reasoning modes and lower API pricing.
Oct 28, 2026
Microsoft brings its multi-model Copilot harness to Foundry
Microsoft has released the shared agent harness behind its Copilot products through Foundry Agent Service.
Oct 27, 2026
Google DeepMind unveils Genie 4 for real-time interactive worlds
Google DeepMind has introduced Genie 4, a natively multimodal world model that generates interactive simulations with visuals, movement, dialogue and ambient audio in real time.
Oct 26, 2026
OpenAI launches GPT-6.1 Terra for lower-cost computer use
OpenAI has launched GPT-6.1 Terra in ChatGPT Work, Codex and its API, positioning the model for visual reasoning and computer-use tasks at lower token prices than GPT-6.1 Sol.
Oct 23, 2026
Moonshot launches Kimi K3.1 Pro and promises open weights by November 4
Moonshot AI has released Kimi K3.1 Pro across its consumer, work, coding and API products, with a faster Highspeed variant and full weights promised by November 4.
Oct 22, 2026
OpenAI releases its Decisions API for production decision systems
OpenAI has moved its Decisions API into general availability, offering bounded classification, scoring and choice outputs on GPT-6 Luna.
Anthropic donates Agent Skills to the Linux Foundation’s agent standards group
Anthropic has transferred its Agent Skills specification to the Agentic AI Foundation under the Linux Foundation.
Oct 21, 2026
Mistral launches Voxtral 3 speech models and a voice-agent SDK
Mistral's new voice generation combines multilingual transcription, speech synthesis and a streaming agent SDK.
Skyfall previews SKY-001 world model for engineering design
Skyfall's new research preview targets structured CAD and building-design workflows.
DeepSeek updates V4.1 Flash with an October 17 checkpoint
DeepSeek's V4.1 Flash-1017 checkpoint is a material update to its existing model, with vendor-reported gains in coding and agent tasks.
Oct 17, 2026
NVIDIA expands its open Nemotron lineup with Super VL and Bolt models
NVIDIA announced Nemotron 3.5 Super VL, a 120-billion-parameter open vision-language model, alongside smaller Bolt models aimed at local agent workloads.
Oct 16, 2026
Meta releases Muse Spark Open weights with an 816B-parameter model
Meta released Muse Spark Open as an open-weight multimodal model with 816 billion total parameters, 42 billion active parameters and a one-million-token context window; its benchmark and desktop-performance statements remain vendor claims.
Mistral launches Timestral, a four-size family of time-series models
Mistral released Timestral, its first time-series foundation model family, with four models from 5M to 209M parameters; its public benchmark leadership claims are based on Mistral-submitted results.
Meta introduces Muse Wonder and Muse Wonder Full for AI agents
Meta introduced Muse Wonder and Muse Wonder Full for agentic work, making Wonder available through its Model API and Muse Code while reserving Wonder Full for Muse Code subscribers.
xAI releases Grok Voice 3.0 for its Voice Agent API
xAI launched Grok Voice 3.0 on October 14 at $0.08 per minute and says it improves speech-agent latency and benchmark performance; those comparative figures remain company-reported.
Google broadens Gemini 4 Argon access after its cyber-first rollout
Google made Gemini 4 Argon globally available to developers and enterprises through its API on October 13 and added access for paid consumer and Antigravity tiers, updating the model's earlier limited cyber rollout.
Oct 8, 2026
Anthropic launches Claude Dashboards and Motion
Anthropic released Claude Dashboards in beta for paid plans and Claude Motion in beta for Team and Enterprise, expanding Claude’s artifact tools from documents and slides into live analytics and animation.
Oct 7, 2026
OpenAI rolls GPT-6 and Intelligent UI into ChatGPT
OpenAI began rolling GPT-6 with Intelligent UI to paid ChatGPT tiers, with Free and Go access starting October 8; the release applies to Chat and does not change the models powering Work or Codex.
Anthropic launches Claude Haiku 5.5 with lower token pricing
Anthropic released Claude Haiku 5.5 across its platform and major clouds, positioning the small model for high-volume work with lower pricing and adjustable effort.
Oct 6, 2026
ElevenLabs launches Architect to build and improve voice agents by conversation
ElevenLabs released ElevenAgents Architect in alpha, a built-in assistant that can analyze agent performance, propose versioned changes, write simulations and build new agent experiences through voice or chat.
Oct 5, 2026
Reflection introduces 501B-parameter Beam model ahead of open-weight release
Reflection says Beam uses 501 billion total parameters, 23 billion active parameters and a 1 million-token context window; weights and technical artifacts are due later in October after final red-teaming.
Oct 2, 2026
Cloudflare releases open Clef decision models and RL fine-tuning
Clef and Clef-flash are hosted on Workers AI with Apache 2.0 weights; Cloudflare also introduced a reinforcement-learning fine-tuning service.
Oct 1, 2026
ElevenLabs launches v4 and Turbo audio-model generation
Eleven v4 and v4 Turbo introduce a new speech-model architecture and broader expressive control.
Sep 29, 2026
OpenAI upgrades GPT-6 Sol with the GPT-6.1 generation
GPT-6.1 Sol is an upgraded Sol generation with new benchmark claims, lower stated API prices, and broader availability across Work and Codex.
OpenAI introduces Dots, its always-on AI agents
OpenAI is rolling out Dots, GPT-6 Astra-powered agents with their own cloud computers and connections to more than 4,000 apps.
Anthropic releases Claude Sonnet 5.5 with faster, lower-cost claims
Anthropic introduced Sonnet 5.5 as the second model in its Claude 5.5 family, reporting faster execution and lower costs than Sonnet 5.
Sep 25, 2026
Microsoft rebuilds Copilot around Home, Code and Autopilot
Microsoft introduced a redesigned Copilot that combines Office work, natural-language app building and a persistent cloud agent.
Oct 2, 2026
Google adds real-time video avatars to Gemini 3.8 Live
Google launched Live Avatar for Gemini 3.8 Live, combining speech, visual input and streaming video avatars for enterprise agents.
Sep 24, 2026
Google launches Gemini 3.8 Flash TTS and Flash-Lite TTS
Google introduced two Gemini 3.8 text-to-speech models for directed voice creation and high-volume generation.
Meta expands its AI-glasses lineup with an audio-only Ray-Ban model
Meta introduced Ray-Ban Meta Audio glasses and said its Muse personal AI agent is coming to the wider glasses lineup.
Sep 22, 2026
Anthropic launches Claude Opus 5.5 with lower operating cost
Anthropic announced Claude Opus 5.5 as a faster, lower-cost flagship model, while pricing, benchmark and safety comparisons remain vendor-reported.
Sep 21, 2026
Light Origins releases Light-O1 preview for whole-body robot intelligence
Light Origins released an open Light-O1 preview for whole-body humanoid actions, with weights and inference materials public but benchmark and scaling claims not independently reproduced.
Sep 20, 2026
Cua releases compact open model for one-pass GUI form selection
Cua released a 706,048-parameter open model for one-pass GUI option selection, with code and training tools available but independently unreproduced benchmark results.
Sep 19, 2026
AI Singapore and NVIDIA release Nemotron-SEA-LION-v4.8 models
AI Singapore and NVIDIA released two open-weight Nemotron-SEA-LION-v4.8 models for Southeast Asian languages; artifacts are public, while benchmark gains remain first-party results.
PrismML releases a compressed 27B open model for local inference
PrismML released an open-weight 27B model compressed to a reported 5.9 GB footprint; its artifact is traceable, while benchmark-retention claims remain vendor-reported.
Sep 14, 2026
Anthropic launches Claude for Financial Advisors with industry connectors
Anthropic launched Claude for Financial Advisors with connectors and workflow skills spanning custody, portfolios, planning and CRM systems.
Sep 12, 2026
Google releases Gemini 3.8 Flash and a guarded cyber variant
Google has introduced Gemini 3.8 Flash for long-horizon coding and agentic workflows, alongside Gemini 3.8 Flash Cyber for trusted defenders.
Sep 11, 2026
OpenAI brings its Codex-style agent harness to developers with the Agents API
OpenAI has put the Agents API into public beta, giving developers a managed harness for long-running cloud agents, tool use, context management, and parallel subagents.
OpenAI brings GPT-Live-1’s full-duplex voice model to the API
OpenAI says GPT-Live-1 is now available in its API as a front-end voice layer that can listen and speak at the same time.