[Dev Catch Up # 114]- Fable 5 is back, OpenAI's GPT 5.6, Claude Sonnet 5, Nano Banana 2 Lite, Gemini Omni Flash, Lore -Open Source Version Control, Hosted X MCP, Nemotron-Labs-TwoTower and more!
Bringing devs up to speed on the latest dev news from the trends including, a bunch of exciting developments and articles
Welcome to the 114th edition of DevShorts, Dev Catch Up.
For those who joined recently or are reading Dev Catch Up for the first time, I write about developer stories and open source, partly based on my work and experience interacting with people all over the globe.
Thanks for reading Dev Shorts! Subscribe for free to receive new posts and support my work.
Some recent issues from Dev Catch up:
Claude Certified Architect Foundations: The Complete Guide - Part 1
Claude Certified Architect Foundations: The Complete Guide - Part 2
Join 8900+ developers to hear stories from Open source and technology.
Must Read
Anthropic is restoring access to Claude Fable 5 after export controls were lifted. Fable 5 is now available globally across Claude Platform. Anthropic has also restored Mythos 5 access for some US organizations. Check Anthropic’s post for more details.
OpenAI has released the GPT 5.6 series. It has three models. Sol as the flagship model, Terra for balanced everyday use, and Luna as the faster and cheaper model. Sol is positioned for harder coding workflows and complex agent tasks. Check OpenAI’s post for more details.
Anthropic has released Claude Sonnet 5. It is built for longer agentic tasks, coding, and tool based workflows. Anthropic says it brings more Opus style agent behavior into the Sonnet tier at a lower price. Check Anthropic’s post for more details.
Google has released Nano Banana 2 Lite and Gemini Omni Flash. Nano Banana 2 Lite is for fast and low cost image generation. Gemini Omni Flash is for video generation and editing. Developers can use both models together to turn images into short videos. Check Google’s post for more details.
OSS Highlight of the Week
This week we are featuring Lore, an open source version control system. It is built for large projects that mix code with large files, like games and entertainment projects. Lore supports lightweight workspaces, fast branching, and file downloads only when needed. Check the GitHub repo for more details.
Good to know
xAI has launched Voice Agent Builder. It is a no code platform for building voice agents on Grok Voice. It is used to build voice agents without setting up the full voice stack from scratch. It also supports telephony, knowledge retrieval, tools, guardrails, MCPs, and observability in one place. Check xAI’s post for more details.
Anthropic has launched Claude Science. Its an AI workbench for scientists. Claude Science is an app that integrates the tools and packages researchers commonly use. It helps with literature review, data analysis, figures, and manuscripts. Check Anthropic’s post for more details.
NotebookLM has added Short Video Overviews. It turns uploaded sources into 60 second mobile videos. This gives users another way to understand the material, apart from text and audio overviews. Check NotebookLM’s post for more details.
Meta has released Brain2Qwerty v2. It turns brain activity into text using non invasive brain recordings. This means the system does not need surgery. Meta is also releasing the training code and the v1 dataset. Check Meta’s post for more details.
Notable FYIs
XAI has announced Hosted X MCP. It is a Streamable HTTP MCP server that lets MCP clients talk to the X API. Developers can use it with their own X account permissions. Check the X Developer Community post for more details.
Claude Code has added an experimental advisor tool. When Claude is unsure or gets stuck during a coding task, it can ask a stronger model for help. It can also use the advisor to review its approach before finishing the task. Check Claude Code docs for more details.
NVIDIA Research has released Nemotron-Labs-TwoTower. It is a diffusion language model built from Nemotron 3 Nano 30B. One part keeps the context, and the other generates tokens in parallel. NVIDIA says this keeps most of the original model quality while making generation faster. Check NVIDIA’s post for more details.
LongCat has released LongCat 2.0, an open source model. It is built for agentic coding. It supports long context work and coding tasks across larger codebases. LongCat says the model was trained and served on Chinese chips. Check LongCat’s post for more details.
Google has open sourced Copybara. The tool is used by Google internally. It transforms and moves source code between repositories. This is useful when teams need to keep private and public repositories synced. Check the GitHub repo for more details.
That’s it from us with this edition. We hope you are going away with a ton of new information. Lastly, share this newsletter with your colleagues and pals if you find it valuable. A subscription to the newsletter will be awesome if you are reading it for the first time.