https://tpsreport.news/news/siri-ai-ios-27-user-complaints TPS — Tokens Per Second en 2026-09-20T18:50:50.754Z New Siri AI in iOS 27 Frustrates Users With Bugs, Slow Responses, and Lock-Screen Failures Siri, Apple, iOS 27, Apple Watch, voice assistant, Private Cloud Compute https://tpsreport.news/news/qwen-image-2-1-7b-open-weight-alibaba TPS — Tokens Per Second en 2026-09-20T16:20:54.187Z Alibaba Releases Qwen-Image-2.1, a 7B-Parameter Open-Weight Image Model It Claims Beats Closed Rivals Qwen, Alibaba, image generation, open-weight models, open source AI, image editing https://tpsreport.news/news/tencent-gander-voice-ai-background-tasks TPS — Tokens Per Second en 2026-09-20T15:51:27.047Z Tencent Unveils Gander, a Voice AI That Keeps Talking While a Separate 'Brain' Handles Background Tasks Tencent, Gander, voice AI, full-duplex conversation, agent systems, Hunyuan Speech, research https://tpsreport.news/news/qwen-image-2-1-7b-text-to-image-editing TPS — Tokens Per Second en 2026-09-20T13:50:47.887Z Alibaba Releases Qwen-Image-2.1, a 7B Unified Text-to-Image and Editing Model Qwen, Alibaba, text-to-image, image editing, open-source, diffusion model, RGBA https://tpsreport.news/news/runway-real-time-video-generation-research TPS — Tokens Per Second en 2026-09-20T12:06:08.723Z Runway Details Research Into Real-Time, Streamed AI Video Generation Runway, video generation, world models, real-time AI, robotics, GWM-1 https://tpsreport.news/news/qwen3-8-omni-flash-undercuts-gemini-pricing TPS — Tokens Per Second en 2026-09-19T14:50:55.287Z Qwen3.8-Omni-Flash Prices Multimodal AI at $0.15/$0.47 per Million Tokens, Undercutting Gemini Flash by 5x Qwen, Alibaba, Gemini Flash, multimodal AI, AI agents, pricing, Qwen3.8-Omni-Flash https://tpsreport.news/news/roboharm-benchmark-gpt-6-astra-claude-fable-safety-failures TPS — Tokens Per Second en 2026-09-19T13:35:51.351Z Robot Safety Benchmark Finds GPT-6 Astra and Claude Fable 5.1 Rarely Refuse Dangerous Commands AI safety, robotics, benchmark, GPT-6 Astra, Claude Fable, MolmoAct2, OpenAI, Anthropic https://tpsreport.news/news/google-deepmind-dream-rsi-search-agents TPS — Tokens Per Second en 2026-09-19T11:21:09.853Z Google DeepMind's Dream-RSI Cuts AI Search Costs by Replaying Past Attempts Instead of Repeating Them Google DeepMind, AI agents, self-improving AI, search algorithms, Gemini, recursive self-improvement, AI research https://tpsreport.news/news/gemini-macos-imessage-integration-1 TPS — Tokens Per Second en 2026-09-19T00:06:32.469Z Gemini for macOS Adds iMessage Integration to Send and Read Messages Gemini, Google, macOS, iMessage, product update, Apple Messages https://tpsreport.news/news/gemini-macos-imessage-integration TPS — Tokens Per Second en 2026-09-19T00:06:14.813Z Gemini for macOS Adds iMessage Integration to Send and Search Messages Gemini, Google, macOS, iMessage, product update, Google DeepMind https://tpsreport.news/news/gemini-hacked-three-companies-google-confirms TPS — Tokens Per Second en 2026-09-19T00:05:51.458Z Google Confirms Gemini Autonomously Breached Three Companies' Systems in May Red-Team Test Gemini, Google DeepMind, AI safety, cybersecurity, red-teaming, autonomous agents, Irregular https://tpsreport.news/news/amazon-sagemaker-inference-2026-launches-review TPS — Tokens Per Second en 2026-09-18T21:05:52.971Z AWS Ships 13 SageMaker Inference Features in 2026, Cutting Startup Latency 51% and GPT-OSS-20B Throughput 2x AWS, Amazon SageMaker, inference, MLOps, GPU infrastructure, OpenAI API compatibility, HyperPod