https://topaiproduct.com/2026/05/11/openai-mrc-protocol-scales-ethernet-to-131000-gpus-and-gets-every-chip-giant-to-sign-on/Top AI Producten2026-05-11T10:07:47+00:00OpenAI MRC Protocol scales Ethernet to 131,000 GPUs — and gets every chip giant to sign onhttps://topaiproduct.com/2026/05/11/xai-ships-grok-connectors-with-8-integrations-and-byo-mcp-support/Top AI Producten2026-05-11T10:05:28+00:00xAI ships Grok Connectors with 8 integrations and BYO MCP supporthttps://topaiproduct.com/2026/05/10/google-remy-geminis-24-7-background-agent-surfaces-8-days-before-i-o/Top AI Producten2026-05-11T04:08:32+00:00Google Remy: Gemini’s 24/7 background agent surfaces 8 days before I/Ohttps://topaiproduct.com/2026/05/10/entire-by-thomas-dohmke-checkpoints-60m-seed-bets-git-wasnt-built-for-ai-agents/Top AI Producten2026-05-11T04:06:07+00:00Entire (by Thomas Dohmke) + Checkpoints: $60M seed bets Git wasn’t built for AI agentshttps://topaiproduct.com/2026/05/10/tencent-hunyuan-hy3-preview-goes-open-source-295b-moe-21b-active-256k-context/Top AI Producten2026-05-10T22:10:07+00:00Tencent Hunyuan Hy3 Preview goes open-source: 295B MoE, 21B active, 256K contexthttps://topaiproduct.com/2026/05/10/prime-intellect-lab-hits-ga-per-token-rl-training-across-14-models/Top AI Producten2026-05-10T22:07:05+00:00Prime Intellect Lab hits GA: per-token RL training across 14 modelshttps://topaiproduct.com/2026/05/10/hkuds-ai-trader-gains-255-github-stars-a-day-hkus-agents-debate-before-they-trade/Top AI Producten2026-05-10T16:11:37+00:00HKUDS AI-Trader gains 255 GitHub stars a day: HKU’s agents debate before they tradehttps://topaiproduct.com/2026/05/10/bytedance-ui-tars-desktop-scores-61-6-on-screenspot-pro-leaving-gpt-4o-and-claude-behind/Top AI Producten2026-05-10T16:07:58+00:00ByteDance UI-TARS-desktop scores 61.6% on ScreenSpot Pro, leaving GPT-4o and Claude behindhttps://topaiproduct.com/2026/05/10/airbyte-agents-context-store-launches-with-50-connectors-and-75-90-token-savings-vs-vendor-mcps/Top AI Producten2026-05-10T10:09:25+00:00Airbyte Agents (Context Store) launches with 50 connectors and 75-90% token savings vs vendor MCPshttps://topaiproduct.com/2026/05/10/google-ships-gemini-api-file-search-multimodal-rag-page-level-citations-and-embedding-2-baked-in/Top AI Producten2026-05-10T10:06:36+00:00Google ships Gemini API File Search Multimodal RAG — page-level citations and Embedding 2 baked inhttps://topaiproduct.com/2026/05/09/anthropic-claude-security-goes-public-beta-after-500-cves-found-in-closed-preview/Top AI Producten2026-05-10T04:15:02+00:00Anthropic Claude Security goes public beta after 500+ CVEs found in closed previewhttps://topaiproduct.com/2026/05/09/anthropic-claude-dreaming-lets-agents-rewrite-their-own-memory-harvey-saw-6x-task-completion/Top AI Producten2026-05-10T04:11:27+00:00Anthropic Claude Dreaming lets agents rewrite their own memory — Harvey saw 6x task completionhttps://topaiproduct.com/2026/05/09/gemini-3-1-flash-lite-hits-ga-0-25-m-input-tokens-2-5x-faster-ttft/Top AI Producten2026-05-09T22:09:39+00:00Gemini 3.1 Flash-Lite hits GA: $0.25/M input tokens, 2.5x faster TTFThttps://topaiproduct.com/2026/05/09/gpt-5-5-cyber-openai-forks-a-security-model-with-looser-guardrails-for-vetted-red-teams/Top AI Producten2026-05-09T22:06:52+00:00GPT-5.5-Cyber: OpenAI forks a security model with looser guardrails for vetted red teamshttps://topaiproduct.com/2026/05/09/agentmemory-rohitg00-hits-95-2-on-longmemeval-27-points-above-mem0/Top AI Producten2026-05-09T16:14:22+00:00agentmemory (rohitg00) hits 95.2% on LongMemEval — 27 points above Mem0https://topaiproduct.com/2026/05/09/subquadratic-subq-claims-1000x-less-compute-at-12m-tokens-researchers-want-receipts/Top AI Producten2026-05-09T16:11:09+00:00Subquadratic SubQ claims 1,000x less compute at 12M tokens — researchers want receipts