Inception kynnir Mercury 3 og segist skáka GPT-6 Luna

22 heimildir
  • Inception hefur gefið út Mercury 3 og segir líkanið standa sig betur en Claude Haiku 5.5 og GPT-6 Luna á sama tíma og það nær 1.068 táknum á sekúndu á Nvidia GPU-skjákortum.
  • Útgáfan frá 19. október er dreifilíkan (e. diffusion model) sem býr til mörg tákn í einu í stað þess að vinna eitt í einu, og boðið er upp á 50% kynningarafslátt.
  • Artificial Analysis, helsta óháða greiningarfyrirtækið á þessu sviði, var ekki með Mercury 3 á lista á útgáfudegi, svo hraðatölur hafa ekki verið staðfestar, samkvæmt AI Model Report.
Heimildir (22)
  1. 1 Inception Launches Mercury 3, Setting a New Pareto Frontier in Speed and Intelligence www.businesswire.com
  2. 2 Inception Releases Mercury 3, Its Flagship Diffusion LLM www.unite.ai
  3. 3 Mercury 3 Ships at 1,068 Tokens/Second and $0.10/M — Without a Single Independent Benchmark — AI Model Report aimodelreport.com
  4. 4 New Mercury 3 model claims 8-fold speed edge over GPT-6 rivals interestingengineering.com
  5. 5 Inception Labs claims Mercury 3 beats Haiku and GPT-6 Luna at 1,068 tokens a second - Startup Fortune startupfortune.com
  6. 6 Inception Mercury 3 review: real speed, limits and who it's for www.eesel.ai
  7. 7 Mercury 3: 1,068 Tokens a Second, and No Benchmarks www.orcarouter.ai
  8. 8 Inception introduces Mercury 3 at 1,068 tokens per second www.siliconreport.com
  9. 9 Mercury 3 vs Reflection Beam: 1,068 Claimed vs 335 Measured www.orcarouter.ai
  10. 10 Mercury 3 from Inception: 1,068 Tokens/s and a 1M-Token Context www.gptunnel.ru
  11. 11 Mercury 3 对比 MiniMax M3:一个数字,一份完整记录 www.orcarouter.ai
  12. 12 Inception Launches Mercury 3 With One-Million-Token Context Window | TokenPost www.tokenpost.com
  13. 13 Inception Mercury 3 Hits 1,068 Tokens/Second at Launch emergent.sh
  14. 14 Inception’s Mercury 3 Undercuts the Speed-Optimized Frontier at 1,068 Tokens per Second forkast.news
  15. 15 Inception's Mercury 3 Scales Diffusion LLMs Over 40x While Holding Generation Speed xenospectrum.com
  16. 16 Inception Releases Mercury 3 Diffusion LLM at 1,068 Tokens per Second www.marktechpost.com
  17. 17 Mercury 3 cresceu 40 vezes e ainda roda a 1.000 tokens por segundo www.orcarouter.ai
  18. 18 Inception’s Mercury 3: 40x Bigger, 8x Faster [2026] shattered.io
  19. 19 Mercury 3 challenges the LLM scaling thesis with diffusion speed news.lavx.hu
  20. 20 New AI model claims to be 8x faster than GPT-6 www.newsbytesapp.com
  21. 21 Anthropic co-founder says the singularity is underway www.therundown.ai
  22. 22 Mercury 3 is fast. Your agent still has to wait for its tools. dev.to