LLM News

Amazon Web Services released Amazon Bedrock Data Automation (BDA), a foundation model-powered service designed to extract and validate structured data from financial documents. The service uses custom blueprints to process bank statements, W-2 tax forms, 1099-B forms, and vendor contracts, offering what AWS claims is industry-leading accuracy at lower cost than using foundation models directly.

Google is rolling out camera-based automations for Gemini for Home, allowing smart home cameras to trigger routines based on visual events like package deliveries or glass breaking. The update includes improved multi-request handling and reduced error rates for the AI assistant that replaced Google Assistant on smart home devices.

AWS has deployed Amazon Bedrock AgentCore to orchestrate more than 20 specialized AI agents across its global sales organization through a unified interface called Field Advisor. The system has processed over 120,000 prompts since launch, delivering a 41% latency reduction compared to previous infrastructure and saving large-scale sales reps up to 2 hours per week on CRM tasks.

Artificial Analysis and IBM Research have released ITBench-AA, the first benchmark evaluating AI models on enterprise Site Reliability Engineering tasks. Claude Opus 4.7 leads at 47%, followed by GPT-5.5 at 46% and Qwen3.7 Max at 42%—all frontier models score below 50% on Kubernetes incident response tasks requiring root-cause diagnosis across complex infrastructure.

OpenAI confirmed it is investigating elevated latency issues affecting both ChatGPT and its API as of May 27, 2026. The company is monitoring a separate issue for FEDRamp users, though a fix has been applied to that specific problem.

ElevenLabs released Music v2, an AI music generation model that can switch genres within a single track and build songs section-by-section. The model, trained on licensed data cleared for commercial use, can transition from opera to heavy metal, handle fast rap, and add sound effects while maintaining coherence.

Google has released Gemini Omni, a multimodal AI video generation tool that accepts text, images, audio, and video as inputs. The first tier, Gemini Omni Flash, includes avatar cloning that creates digital versions of users and incorporates physics modeling for realistic motion.

Google is bundling Health Premium into its AI Pro subscription tier at $19.99 per month. The service, which includes a Gemini-powered health coach and adaptive fitness plans, was previously available as a standalone product at $9.99 monthly or $99.99 annually.

Meta is rolling out document upload support to Meta AI in WhatsApp for iOS beta users, according to WABetaInfo. The feature, already available to Android beta testers, allows users to share PDFs and spreadsheets directly from the in-chat attachment menu instead of relying on screenshots.

GitHub has added organization-level model targeting to Copilot Enterprise. The feature allows enterprise owners to control which Copilot models are available to specific organizations within their deployment, replacing the previous all-or-nothing approach.

Microsoft released Lens, a 3.8-parameter foundational text-to-image model trained on Lens-800M, an 800 million image-text corpus with GPT-4.1 captions. The model uses a 48-block MMDiT denoiser with FLUX.2 latents and supports generation up to 1440×1440 resolution across aspect ratios from 1:2 to 2:1.

Cline, the AI coding assistant VS Code extension, released version 3.85.0 on May 25, 2025, adding support for DeepSeek V4 Flash and Pro models, Gemini 3.5 Flash across multiple providers, and GPT-5.5 through SAP AI Core. The update also fixes Vertex AI global endpoint handling for Claude models.

Microsoft has released Lens-Turbo, a 3.8B-parameter foundational text-to-image model designed for efficient training and fast generation. The model was trained on Lens-800M, an 800 million image-text corpus with GPT-4.1 captions, and supports resolutions up to 1440×1440 with 4-step distilled inference.

Apple plans to significantly improve the visual quality of its on-device image generation models for Genmoji and Image Playground in iOS 27, according to Bloomberg's Mark Gurman. The update will also add support for third-party AI image generation models beyond OpenAI's ChatGPT.

Stability AI has released Stable Audio 3 Medium, a 2 billion parameter latent diffusion model capable of generating variable-length audio up to 380 seconds. The model generates music and sound effects in less than 2 seconds on an H200 GPU, trained on 1.28 million licensed and Creative Commons audio recordings.

Google has expanded Gemini Live's Connected Apps integration on Android, adding support for 15 new services including YouTube Music, Spotify, Home controls, Flights, Hotels, Workspace, and Utilities. The update includes a redesigned floating interface that allows users to switch between text and voice conversations.

DeepSeek has permanently reduced pricing for its V4 Pro model by 75%, bringing input token costs down to $0.003625 per million tokens from $0.0145. The move makes permanent a promotional discount that was set to expire May 31, 2026.

Anthropic's unreleased Claude Mythos Preview model has discovered more than 10,000 vulnerabilities across partner organizations in its first month of deployment through Project Glasswing. The company reports partners are finding bugs at 10x their previous rate, with Cloudflare discovering 2,000 bugs and Mozilla finding 271 Firefox vulnerabilities — 10x more than with previous Claude models.

Tencent released Hy-MT2, a family of multilingual translation models available in 1.8B, 7B, and 30B-A3B (MoE) sizes. All models support translation among 33 languages and follow translation instructions in multiple languages. The 1.8B model can be compressed to 440MB using 1.25-bit AngelSlim quantization.

NVIDIA has released Nemotron-Labs Diffusion, a family of diffusion language models at 3B, 8B, and 14B scales that generate multiple tokens in parallel rather than one at a time. The 8B model achieves 6.4× higher tokens per forward pass than autoregressive models in self-speculation mode while maintaining comparable accuracy.