OpenAI is preparing the GPTâ5.1 family, Google readies a November preview of Gemini 3 Pro with a 1Mâtoken context, and an updated Nano Banana (GEMPIX 2) image model is on deck.
Google is preparing to launch GEMPIX 2, an updated Nano Banana image generation model in its Gemini lineup, aimed at creative professionals and expected to offer faster, high-performance outputs.
Apple plans to pay $1B yearly to use Googleâs 1.2T parameter Gemini AI model to upgrade Siriâs core functions in iOS 26.4, marking a major but temporary shift as Apple works on its own competing AI system.
Gemini 3 Pro, Googleâs next-gen multimodal AI model with a 1M token context window, is previewing in VertexAI and expected to launch broadly in December, potentially alongside Nano Banana 2.
Anthropic expects $70B revenue and $17B cash flow by 2028, fueled by rapid B2B growth, major partnerships, and Claudeâs expanding product suite, with gross margins projected to reach 77%.
Shopify reported 7x AI-driven traffic growth and 11x AI-attributed purchases since January, calling AI core to its strategy and partnering with OpenAI, Microsoft, and Perplexity to lead in âagentic commerce.â
Anthropic signed a major deal with Cognizant to deploy Claude AI to 350,000 employees and co-sell industry-specific AI solutions, deepening its enterprise push and boosting Cognizantâs AI capabilities.
Kimi K2 Thinking is a powerful open-source agent model excelling in reasoning, coding, and search, achieving 44.9% on HLE, 60.2% on BrowseComp, and 71.3% on SWE-Bench Verified with up to 300 autonomous tool calls.
Cursorâs semantic search boosts coding agent accuracy up to 23.5%, improves code retention, and reduces user correctionsâespecially on large codebasesâby using a custom-trained embedding model on real agent behavior.
The Gemini API now includes the File Search Tool. It is a fully managed RAG system that simplifies data grounding for more accurate responses.
Autonomous AI agents that act just like real hackers - they run your code dynamically, find vulnerabilities, and validate them through actual proof-of-concepts.
An open-source, code-first Go toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.
These traces appear in the code for Enterprise RBAC settings, and the release date is set to "November 24".
It's a powerful, general-purpose model that excels across real-world tasks, with standout performance in coding, tool calling, and instruction following.
When augmented with tool use and scaled test-time compute, it achieves 100% on challenging reasoning benchmarks like AIME 2025 and HMMT.
Users will be able to create virtual environments and run Grok Code tasks directly from the web.
Three updates including GPT-5-Codex-Mini (a more compact and cost-efficient version of GPT-5-Codex ), 50% higher rate limits, and priority processing for ChatGPT Pro and Enterprise.
Just tag Codex to answer questions, fix bugs, and explore new ideas, right from a team channel or thread.
MCP enables tool-rich AI agents, but code execution reduces token costs by 98.7%, improves control flow, and preserves privacy by letting agents fetch tools and process data via code instead of context.
A curated weekly roundup of AI news, tools, and breakthroughs to stay ahead of the curve in a fast-moving industry.
Curated by devs, for devs. Helping 400+ developers stay ahead of the AI curve.