Skip to content

Category

New models

58 events 12 over 7 days ↓ 8 % 24 sources report highest score 88

Importance: important New models ✓ official

RadixArk releases open-weight model DaisyRose-27B for agentic tasks

RadixArk has released the DaisyRose-27B model under the Apache 2.0 license, a fine-tuned version of the Qwen3.8-27B model focused on long agentic tasks. On the Terminal Bench 4.1 benchmark it improved pass@1 from 5.6% to 12.0%, and other coding and instruction-following tests also showed improvement.

Importance: important OpenAI New models ✓ official

GitHub Copilot rolls out OpenAI's GPT-6.2 Astra and GPT-6.2 Sol models

GitHub Copilot now offers OpenAI's GPT-6.2 Astra and GPT-6.2 Sol models for agentic coding and code review. According to GitHub, in tests they caught more issues than the older GPT-6 models while consuming fewer tokens. Available for Pro+, Max, Business, and Enterprise, with a gradual rollout.

Importance: important NVIDIA New models ✓ official

NVIDIA introduces Nemotron for Science initiative for scientific agents

NVIDIA has announced the Nemotron for Science initiative, which combines the open Nemotron 3.5 Super VL model, a new open chemistry dataset, and post-training recipes for building scientific agents.

Importance: important Anthropic New models update ✓ 2

Anthropic expands Claude's office tools with connections to Snowflake and Salesforce

Anthropic has clarified the recently introduced suite of office tools in the Claude product: Claude Data now connects to Snowflake and Salesforce, third-party connectors now also include Salesforce alongside Adobe and Atlassian, Sheets supports formulas, and Whiteboards summarizes ideas.

New: Claude Data specifically connects to Snowflake and Salesforce; Interactive Connectors integrate Salesforce (alongside Adobe and Atlassian); Claude Sheets supports formulas and multiple tabs; Claude Whiteboards enables summarization of ideas; Claude Meetings is being prepared for recording and note-taking from calls

Importance: important Anthropic New models update ✓ official

Claude Fable 5.2 model in GitHub Copilot: Business and Enterprise administrators must manually enable data retention

GitHub clarified that the data retention policy for the Claude Fable 5.2 model is disabled by default in the Copilot Business and Enterprise plans, and administrators must enable it manually. The model has been available since 9 November 2026 in all paid Copilot plans.

New: Claude Fable 5.2 data retention policy is disabled by default in Copilot Business and Enterprise; Administrators of the Business and Enterprise versions must manually enable the policy; Specific availability date: 2026-11-09

Importance: important OpenAI New models update ✓ 2

OpenAI launches animated GIF generation in ChatGPT and Codex using the Images 2.6 model

OpenAI is expanding Images 2.6 with animated GIF generation in ChatGPT and Codex, and has released the GPT-Image-2.6 Flare and Sunburst API models featuring SVG generation and markup-guided editing. The company reports more than 4 billion images generated weekly.

New: The rollout of GIF generation is also extending to Codex, not just ChatGPT

Importance: important Mistral New models update ✓ official

Mistral AI launched the Mistral Studio platform for deploying AI agents alongside the release of the Mistral Large 4 model

Alongside the release of the Mistral Large 4 model weights, Mistral AI launched the Mistral Studio platform for deploying agents, workflows, and applications in the cloud, VPC, and on-premises. Studio is already used in production by ABANCA, BNP Paribas, Accenture, and other companies.

New: Mistral Studio is a new platform released alongside Mistral Large 4, focused on enterprise AI; Studio includes managed agents, workflows, enterprise context, and agent observability; Studio supports deployment on the cloud, VPC, and on-premises without dependence on a single provider; the Mistral MCP server enables integration with your own systems; ABANCA, BNP Paribas, Accenture, and 7 other companies already use Studio in production

Importance: important NVIDIA New models update ✓ official

NVIDIA released the Granary v2 dataset with 844 thousand hours of audio for Nemotron 3.5 Transcribe

Nemotron 3.5 Transcribe has 30 billion parameters, approximately 4 billion of which are active, and runs on a single GPU according to NVIDIA. The release is accompanied by the Granary v2 dataset with 844 thousand hours of audio and tools for domain adaptation.

New: The model has 30 billion parameters, with approximately 4 billion active in a single step.; The model can run on a single GPU.; The release includes the Granary v2 dataset with 844 000 hours of audio.; The release includes tools for adapting the model to specialized domains.

Importance: major Google New models update ✓ 4

Google has launched production access to the Genie 4 model for Google AI Ultra subscribers

Google has launched production access to the Genie 4 model (previously only a research preview) for Google AI Ultra subscribers in the Project Genie app, with world edits during simulation and real-time audio. It has also updated the Genie 3 model and added a community gallery.

New: Genie 3 will be updated to support sessions of up to 3 minutes (compared to 1 minute originally); Genie 3 gains 360° panorama and chainable worlds features; The community gallery includes SynthID watermarks and safety checks; A new AI agent enables building worlds from spoken natural-language descriptions; Genie 4 enables real-time scene edits via text prompts (weather, characters, plot)

Importance: major OpenAI New models update ✓ 4

GPT-6.1 Terra has half the API input and output rates of GPT-6.1 Sol

GPT-6.1 Terra costs 1 USD per million input tokens and 5 USD per million output tokens through the API, half the rates for GPT-6.1 Sol. According to OpenAI, it focuses on office tasks, visual reasoning and automated software operation.

New: GPT-6.1 Sol costs 2 USD per million input tokens.; GPT-6.1 Sol costs 10 USD per million output tokens.; GPT-6.1 Terra has half the input and output token prices of GPT-6.1 Sol.

Importance: important New models update ✓ 2

Flint targets agentic coding with optimizations for Apple and Nvidia hardware

According to Recursive, Flint is optimized for fast local agentic coding on Apple and Nvidia hardware. It is available for free under the Apache 2.0 license; the published performance test results come from the company.

New: Flint is optimized to run locally at high speed on Apple and Nvidia hardware.; Flint is also intended for agentic coding.

Importance: important New models ✓ official

IBM releases granite-speech-5.0-2b for multilingual transcription and speech translation

IBM has released granite-speech-5.0-2b, a model with 2 billion parameters for transcription in six languages and speech translation. It offers prioritization of up to 2 000 terms and an Apache 2.0 license. According to the company, it handles noise and longer translations better.

Importance: important Mistral New models update ✓ official

Transcription with Voxtral 3 models starts at 0.0035 USD per minute, SDK uses LiveKit

Transcription pricing in the Voxtral 3 model family starts at 0.0035 USD per minute. The SDK for voice agents uses LiveKit, and according to Mistral AI, speech synthesis adapts emotion, pace and emphasis to the context of the conversation.

New: Transcription pricing starts at 0.0035 USD per minute.; The SDK is based on LiveKit.; Speech synthesis supports emotional adaptation.

Importance: major Google New models update ✓ official

Google has released Gemini 4 Argon globally as a frontier reasoning model for coding, finance, and law

Google has expanded the availability of the Gemini 4 Argon model from paid developer programs to a global release as a frontier reasoning model for coding, financial analysis, and legal workflows; it continues to offer the robotics variant separately.

New: The model has been released globally, accessible via Google AI Studio, Gemini API, Google Cloud, Search, and the Gemini app; Gemini 4 Argon is classified as a frontier reasoning model; Focused on complex coding, financial analysis, and legal workflows in addition to robotics; Specific availability expanded from paid developer programs to global distribution; Robotics is a special version with physical reasoning, not the primary focus

Importance: important NVIDIA New models update ✓ 3

Nvidia's Nemotron 3.5 Super VL model is now available in Amazon SageMaker JumpStart

Nvidia's Nemotron 3.5 Super VL model is now available in Amazon SageMaker JumpStart, in NVFP4 (75 GB) and BF16 (249 GB) variants with a context of up to 1 million tokens. The model, with 120 billion parameters, was previously released as open for enterprise agents and document analysis.

New: The model is now available in Amazon SageMaker JumpStart; Supports context length of up to 1 million tokens; There are two variants: NVFP4 (75 GB, quantized) and BF16 (249 GB, full precision); The model includes a Multi-Token Prediction head for speculative decoding; Specific use cases include document intelligence, GUI understanding for computer-use agents, and video question answering

Importance: important Meta New models update ✓ official

Muse Spark Open weights are available on Hugging Face, with conditions set by the license

The weights for Muse Spark Open from Meta are available on Hugging Face. The free license requires the attribution “Built with Muse Spark Open”; API providers with annual revenue above 20 million dollars need a separate agreement.

New: The weights are available on Hugging Face.; According to Meta, the model runs on two Mac Studio computers with 512 GB of unified memory each.; The license requires the attribution “Built with Muse Spark Open”.; API providers with annual revenue above 20 million dollars need a separate agreement.

Importance: important New models ✓ official

FLUX.3 Image Klein is available in the Diffusers library for image generation and editing

The FLUX.3 Image Klein model from Black Forest Labs is available in the Diffusers library with two variants of the weights. It supports text-to-image generation, editing with multiple reference images, and control over the placement of elements using bounding boxes.