https://tpsreport.news/news/rabbit-os3-ai-agent-no-r1-required TPS — Tokens Per Second en 2026-09-22T21:05:45.650Z Rabbit Launches OS3, a Cloud-Based AI Agent That Runs Without the R1 Device Rabbit, AI agent, OS3, desktop AI, R1 https://tpsreport.news/news/gpt-6-prompt-caching-update TPS — Tokens Per Second en 2026-09-22T20:21:49.322Z OpenAI Rolls Out Improved Prompt Caching for GPT-6 OpenAI, GPT-6, prompt caching, API, changelog, latency, cost optimization https://tpsreport.news/news/gpt-6-sol-luna-price-cut-flat-performance TPS — Tokens Per Second en 2026-09-22T20:21:27.290Z OpenAI Cuts GPT-6 Sol and Luna Prices in Half, but Independent Benchmarks Show Flat Performance OpenAI, GPT-6, pricing, Anthropic, Claude Opus 5, benchmarks, Artificial Analysis https://tpsreport.news/news/gpt-6-prompt-caching-improvements TPS — Tokens Per Second en 2026-09-22T20:20:50.484Z OpenAI Rolls Out Improved Prompt Caching for GPT-6 OpenAI, GPT-6, prompt caching, API, changelog, latency, cost optimization https://tpsreport.news/news/claude-opus-5-5-gpt-6-sol-luna-price-cuts TPS — Tokens Per Second en 2026-09-22T19:51:25.077Z Anthropic and OpenAI Cut Prices With Claude Opus 5.5, GPT-6 Sol and GPT-6 Luna Anthropic, OpenAI, Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, AI pricing, AI safety https://tpsreport.news/news/gpt-6-sol-pro-luna-openrouter-listings TPS — Tokens Per Second en 2026-09-22T19:36:11.859Z OpenRouter Listings Surface for Three Unannounced OpenAI Models: GPT-6 Sol Pro, Luna, and Luna Pro OpenAI, GPT-6, OpenRouter, unverified, rumor, model release, roundup, trending https://tpsreport.news/news/qwen3-8-omni-flash-audio-video-agentic-model TPS — Tokens Per Second en 2026-09-22T19:35:49.639Z Qwen3.8 Omni Flash: Alibaba's First Agentic Omni-Modal Model Adds Native Audio-Video Understanding, 1M Context Qwen, Alibaba, omni-modal, multimodal AI, agentic AI, audio-video model https://tpsreport.news/news/gpt-6-sol-pro-luna-luna-pro-openrouter-listings TPS — Tokens Per Second en 2026-09-22T19:21:55.177Z Three Unverified 'GPT-6' Model Listings Appear on OpenRouter: Sol Pro, Luna, and Luna Pro OpenAI, GPT-6, OpenRouter, unconfirmed, rumor, roundup, trending https://tpsreport.news/news/claude-opus-5-5-gpt-6-sol-luna-pricing TPS — Tokens Per Second en 2026-09-22T19:21:06.053Z Anthropic Ships Claude Opus 5.5, OpenAI Launches GPT-6 Sol and Luna — All Cheaper Than Predecessors Anthropic, OpenAI, Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, AI pricing, model release https://tpsreport.news/news/gpt-6-luna-pro-reasoning-mode TPS — Tokens Per Second en 2026-09-22T18:37:43.799Z OpenAI Lists GPT-6 Luna Pro: A High-Reasoning Mode for Its Budget GPT-6 Model, Not a New Checkpoint OpenAI, GPT-6, reasoning models, OpenRouter, pricing https://tpsreport.news/news/openai-gpt-6-luna-release TPS — Tokens Per Second en 2026-09-22T18:37:17.959Z OpenAI Launches GPT-6 Luna: Fast, Low-Cost Model With 1.1M Context Window OpenAI, GPT-6, GPT-6 Luna, model release, pricing, context window https://tpsreport.news/news/gpt-6-sol-pro-reasoning-mode TPS — Tokens Per Second en 2026-09-22T18:36:16.889Z OpenAI Launches GPT-6 Sol Pro, a High-Reasoning Mode for Its Mid-Tier Model at $2/$10 per 1M Tokens OpenAI, GPT-6, GPT-6 Sol, reasoning models, API pricing, OpenRouter https://tpsreport.news/news/openai-python-v3-18-0-gpt-6-sol-luna-identifiers TPS — Tokens Per Second en 2026-09-22T18:35:47.942Z OpenAI Python SDK v3.18.0 Adds Identifiers for 'GPT-6 Sol' and 'GPT-6 Luna' OpenAI, openai-python, SDK, GPT-6, changelog https://tpsreport.news/news/gpt-6-sol-luna-openai-release TPS — Tokens Per Second en 2026-09-22T18:21:50.721Z OpenAI Launches GPT-6 Sol and GPT-6 Luna, Cutting API Prices 50% Versus GPT-5.6 OpenAI, GPT-6, ChatGPT, Codex, model release, API pricing, coding agents https://tpsreport.news/news/gpt-6-sol-luna-amazon-bedrock TPS — Tokens Per Second en 2026-09-22T18:20:57.038Z OpenAI's GPT-6 Sol and GPT-6 Luna Launch on Amazon Bedrock, Priced Below GPT-5.6 GPT-6, OpenAI, Amazon Bedrock, AWS, model release, enterprise AI https://tpsreport.news/news/openai-gpt-6-sol-luna-launch TPS — Tokens Per Second en 2026-09-22T18:06:25.059Z OpenAI Releases GPT-6 Sol and Luna at Half the Price of Predecessors OpenAI, GPT-6, GPT-6 Sol, GPT-6 Luna, ChatGPT, API pricing, Anthropic https://tpsreport.news/news/claude-opus-5-5-available-aws-bedrock TPS — Tokens Per Second en 2026-09-22T17:36:04.991Z Claude Opus 5.5 Launches on Amazon Bedrock, Anthropic's First Model in New 5.5 Family Claude Opus 5.5, Anthropic, Amazon Bedrock, AWS, model release, agentic AI https://tpsreport.news/news/anthropic-claude-opus-5-5-release TPS — Tokens Per Second en 2026-09-22T16:51:41.896Z Anthropic Releases Claude Opus 5.5, Cuts Pricing 20% and Claims Frontier Coding Lead Anthropic, Claude, Claude Opus 5.5, AI model release, coding AI, agentic AI https://tpsreport.news/news/claude-opus-5-5-40-percent-cheaper-fable-5-1 TPS — Tokens Per Second en 2026-09-22T16:51:07.106Z Anthropic Releases Claude Opus 5.5, Cuts Costs 40% While Matching Rival Fable 5.1 Claude Opus 5.5, Anthropic, model release, pricing, benchmarks, AI safety, distillation https://tpsreport.news/news/anthropic-python-sdk-v1-8-0-claude-opus-5-5 TPS — Tokens Per Second en 2026-09-22T16:37:33.704Z Anthropic SDK for Python v1.8.0 Adds Support for Claude Opus 5.5, Fixes Streaming Crash Anthropic, SDK, Python, Claude, changelog, API, MCP https://tpsreport.news/news/claude-opus-5-5-cybersecurity-safeguards TPS — Tokens Per Second en 2026-09-22T16:37:07.510Z Anthropic Releases Claude Opus 5.5 With Tighter Cybersecurity Safeguards After Rogue AI Incidents Anthropic, Claude, Claude Opus 5.5, AI Safety, Cybersecurity, Model Release https://tpsreport.news/news/anthropic-opus-5-5-lower-pricing TPS — Tokens Per Second en 2026-09-22T16:36:33.853Z Anthropic Releases Claude Opus 5.5, Cuts Output Pricing to $20 per Million Tokens Anthropic, Claude, Opus 5.5, AI pricing, AI safety https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-open-model-anthropic-claude TPS — Tokens Per Second en 2026-09-22T11:51:03.644Z Xiaomi's MiMo-V2.6-Pro Tops Open Model Rankings at $0.13 Per Task, But Anthropic Says It Used Claude to Get There Xiaomi, MiMo, Anthropic, Claude, open models, reinforcement learning, model distillation, Artificial Analysis https://tpsreport.news/news/openai-internal-model-100-math-problems TPS — Tokens Per Second en 2026-09-22T11:35:47.112Z OpenAI Claims Unnamed Internal Model Solved 100+ Open Math Problems After One Month of Training OpenAI, mathematics, AI research, Millennium Prize Problems, Timothy Gowers, AI safety https://tpsreport.news/news/anthropic-biology-lab-claude-robots-drug-experiments TPS — Tokens Per Second en 2026-09-22T09:35:46.312Z Anthropic Builds Physical Biology Lab Where Claude Directs Robots to Run Drug Experiments Anthropic, Claude, drug discovery, biology lab, robotics, life sciences, AI in healthcare https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-1t-a42b TPS — Tokens Per Second en 2026-09-22T06:51:00.344Z Xiaomi's MiMo-V2.6-Pro Becomes Top Open-Weights Model, Trained for $3M According to Xiaomi Xiaomi, open weights, MiMo, MoE, reinforcement learning, Chinese AI labs, Artificial Analysis https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-rl-1m-context TPS — Tokens Per Second en 2026-09-22T03:06:34.563Z Xiaomi Releases MiMo-V2.6-Pro-RL, a 1.02T-Parameter Omnimodal Model with 1M-Token Context Xiaomi, MiMo, open weights, mixture-of-experts, reinforcement learning, multimodal, long context, agentic AI https://tpsreport.news/news/xiaomi-mimo-v2-6-flash-rl-1m-context TPS — Tokens Per Second en 2026-09-22T00:51:08.921Z Xiaomi Releases MiMo-V2.6-Flash-RL, a 309B-Parameter MoE Model with 1M-Token Context and Native Omnimodal Support Xiaomi, MiMo, open-source, MoE, long-context, omnimodal, reinforcement-learning, agentic-AI https://tpsreport.news/news/jev-typesafe-ai-decision-model TPS — Tokens Per Second en 2026-09-21T23:20:59.048Z TypeSafe AI Launches Jev, a 'Decision Model' That Outputs Only Numbers, Priced at $0.042/M Input Tokens Jev, TypeSafe AI, decision models, System One, classification models, LLM pricing, AI bias https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-flash-ultraspeed-roundup TPS — Tokens Per Second en 2026-09-21T20:38:07.521Z Xiaomi Lists Three New MiMo-V2.6 Models on OpenRouter: Pro, Flash, and Pro-UltraSpeed Xiaomi, MiMo, OpenRouter, model release, China AI, roundup, trending https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-ultraspeed TPS — Tokens Per Second en 2026-09-21T20:37:29.195Z Xiaomi Launches MiMo-V2.6-Pro-UltraSpeed: Same Quality, 10x Faster Output Xiaomi, MiMo, model release, multimodal, agentic AI, long context, inference speed https://tpsreport.news/news/xiaomi-mimo-v2-6-flash-1m-context TPS — Tokens Per Second en 2026-09-21T20:36:39.565Z Xiaomi Releases MiMo-V2.6-Flash: Open-Source MoE Model with 1M-Token Context, $0.14/$0.28 per 1M Tokens Xiaomi, MiMo, open-source, Mixture-of-Experts, agentic AI, long-context, multimodal https://tpsreport.news/news/xiaomi-mimo-v2-6-pro-1m-context TPS — Tokens Per Second en 2026-09-21T20:35:53.120Z Xiaomi Launches MiMo-V2.6-Pro, a 1T+ Parameter Model with 1M-Token Context Xiaomi, MiMo-V2.6-Pro, model release, multimodal, long context, agentic AI, OpenRouter https://tpsreport.news/news/yandex-aliceai-foundation-80b-a3b-base TPS — Tokens Per Second en 2026-09-21T19:52:22.094Z Yandex Releases AliceAI-Foundation-80B-A3B-Base, an 80B-Parameter MoE Model with 262K Context Yandex, open-source, MoE, base-model, Russian-language, AliceAI https://tpsreport.news/news/vercel-ai-sdk-xai-grok-4-7-model-id TPS — Tokens Per Second en 2026-09-21T19:51:11.173Z Vercel AI SDK Adds Grok 4.7 Model ID Support in xai Provider Patch 3.0.134 xAI, Grok, Vercel AI SDK, changelog, SDK update https://tpsreport.news/news/ai-sdk-xai-2-0-95-grok-4-7-model-id TPS — Tokens Per Second en 2026-09-21T19:50:51.878Z Vercel AI SDK Adds Support for Grok 4.7 Model ID in xAI Provider Package xAI, Grok, Vercel AI SDK, changelog, developer tools https://tpsreport.news/news/ai-sdk-xai-5-0-5-grok-4-7-model-id TPS — Tokens Per Second en 2026-09-21T19:36:14.481Z Vercel AI SDK Patch Adds Grok 4.7 Model ID Support Vercel, AI SDK, xAI, Grok, changelog, developer tools https://tpsreport.news/news/meta-muse-ai-agent-tops-chatgpt-downloads TPS — Tokens Per Second en 2026-09-21T19:20:52.192Z Meta's Muse AI Agent App Hits 730,000 Downloads, Overtakes ChatGPT on iOS Charts Meta, Muse, AI agents, ChatGPT, app downloads, Mark Zuckerberg https://tpsreport.news/news/grok-4-6-amazon-bedrock-availability TPS — Tokens Per Second en 2026-09-21T18:50:56.907Z xAI's Grok 4.6 Launches on Amazon Bedrock With 500K Context and Cross-Region Inference Grok 4.6, xAI, Amazon Bedrock, AWS, model availability, agentic AI, Converse API https://tpsreport.news/news/grok-4-7-pricing-benchmarks-claude-gpt-6 TPS — Tokens Per Second en 2026-09-21T17:05:53.611Z xAI Releases Grok 4.7 at $2/$6 per Million Tokens, Trails Claude and GPT-6 on Benchmarks Grok 4.7, xAI, Elon Musk, benchmarks, pricing, Claude, GPT-6, agentic coding https://tpsreport.news/news/cactus-compute-needle-3-on-device-model TPS — Tokens Per Second en 2026-09-21T16:51:07.023Z Cactus Compute Releases Needle 3, a Sub-30MB On-Device Model for Tool Calling and Extraction Cactus Compute, Needle 3, on-device AI, edge AI, tool calling, structured extraction, small language models, open weights https://tpsreport.news/news/grok-4-7-500k-context-release TPS — Tokens Per Second en 2026-09-21T16:36:11.262Z xAI Ships Grok 4.7, Cuts Price to $1.60/$4.80 per 1M Tokens With 500K Context Grok 4.7, xAI, OpenRouter, model release, long context, agentic AI, coding models https://tpsreport.news/news/balance-of-power-open-weight-models-china-us TPS — Tokens Per Second en 2026-09-21T12:06:16.496Z Chinese Open-Weight Models Now Lead US Rivals by 2-6 Months, Congressional Briefing Shows open-weight models, China AI, policy, Qwen, GLM, Kimi K3, benchmarks, AI competition https://tpsreport.news/news/chinese-open-weight-models-lead-2026-analysis TPS — Tokens Per Second en 2026-09-21T12:05:51.957Z Chinese Open-Weight Models Now Lead US Rivals by Wide Margin, Interconnects Analysis Finds open-weight models, China AI, GLM-5.3, Kimi K3, Qwen, DeepSeek, AI policy, Hugging Face