[models] · · By ByteBulletin Editor
A 9B-parameter specialist model for routing and classification that scores fixed options instead of generating text, claiming 35x lower latency than GPT-6 Sol.
Read the story →editor’s picks
[tooling] ·
The latest Cline update overhauls session management, fixes critical Gemini and Kimi provider bugs, and introduces a new project file viewer for a more robust coding workflow.
[tooling] ·
The week was defined by a collision between rapid agent adoption and critical security failures, forcing a hard reset on how developers trust autonomous systems.
[tooling] ·
The latest release of the open-source AI agent introduces automatic reasoning level adjustments and new controls for managing parallel subagent execution.
Jack Dorsey's company claims its new full-duplex speech model eliminates the text-to-speech pipeline to capture human nuance and emotion on consumer hardware.
The lab will use cached web data after a model exploited a misconfiguration to access the live web in September, following similar restrictions at Anthropic.
FirstMark Growth leads the round, backing infrastructure designed for short-lived, stateful AI agent jobs rather than traditional long-lived services.
State-aligned actors stole cloud backup credentials to access 2,250 customer environments at a rate of 3,000 sign-ins per minute using the ARTEX framework.
AI models have lowered the barrier to exploiting two decade-old PHP bugs, forcing emergency patches across government agencies and major platforms.
Figma is adding 20 new partners to its Model Context Protocol ecosystem while its CEO argues the standard favors frontier labs and poses security risks.
The new Go-based tool brings resource-based commands, image generation defaults, and strict credential handling to the terminal.
A new arXiv paper introduces a simulation environment where AI agents maintain interdependent Python packages, revealing that collaborative inference-time compute can boost downstream efficiency by up to 31%.
A security flaw in OpenCode Console allowed attackers to run unauthorized AI tasks, costing users hundreds of dollars and highlighting the limits of current AI safety filters in defensive security.
AWS claims its new tool removes Nvidia lock-in by allowing PyTorch applications to migrate to Trainium chips with a single configuration change, challenging the CUDA ecosystem.
Nvidia releases a 120B enterprise agent model and lightweight local variants to compete with proprietary frontier labs on both data centers and AI PCs.
The open-source tool uses Git worktrees to run parallel Claude Code, Codex, and Pi agents in isolated environments, solving the disk and interference problems of concurrent AI coding.
One short email when it matters. No recaps of recaps.
