home.social

#chinesechips — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #chinesechips, aggregated by home.social.

fetched live
  1. #Meituan open-sourced #LongCat 2.0, a 1.6T parameter Mixture-of-Experts model trained entirely on #Chinesechips. The model, which has been leading #OpenRouter, offers a 1-million-token context window and is available under a permissive #MITlicence. LongCat-2.0 excels in software engineering tasks, outperforming GPT-5.5 on the SWE-bench Pro benchmark. venturebeat.com/technology/mei #tech #media #news

  2. #Meituan open-sourced #LongCat 2.0, a 1.6T parameter Mixture-of-Experts model trained entirely on #Chinesechips. The model, which has been leading #OpenRouter, offers a 1-million-token context window and is available under a permissive #MITlicence. LongCat-2.0 excels in software engineering tasks, outperforming GPT-5.5 on the SWE-bench Pro benchmark. venturebeat.com/technology/mei #tech #media #news

  3. #Meituan open-sourced #LongCat 2.0, a 1.6T parameter Mixture-of-Experts model trained entirely on #Chinesechips. The model, which has been leading #OpenRouter, offers a 1-million-token context window and is available under a permissive #MITlicence. LongCat-2.0 excels in software engineering tasks, outperforming GPT-5.5 on the SWE-bench Pro benchmark. venturebeat.com/technology/mei #tech #media #news

  4. #Meituan open-sourced #LongCat 2.0, a 1.6T parameter Mixture-of-Experts model trained entirely on #Chinesechips. The model, which has been leading #OpenRouter, offers a 1-million-token context window and is available under a permissive #MITlicence. LongCat-2.0 excels in software engineering tasks, outperforming GPT-5.5 on the SWE-bench Pro benchmark. venturebeat.com/technology/mei #tech #media #news

  5. #Meituan open-sourced #LongCat 2.0, a 1.6T parameter Mixture-of-Experts model trained entirely on #Chinesechips. The model, which has been leading #OpenRouter, offers a 1-million-token context window and is available under a permissive #MITlicence. LongCat-2.0 excels in software engineering tasks, outperforming GPT-5.5 on the SWE-bench Pro benchmark. venturebeat.com/technology/mei #tech #media #news

  6. #Meituan trained a 1.6 trillion parameter AI model, #LongCat 2.0, entirely on #Chinesechips, demonstrating #China’s capability to train #largescalemodels without #Nvidia hardware. While LongCat-2.0 outperforms some Western models on specific benchmarks, it falls short on others. the-decoder.com/meituans-longc #AIagent #AI #ML #NLP #LLM #GenAI

  7. #Meituan trained a 1.6 trillion parameter AI model, #LongCat 2.0, entirely on #Chinesechips, demonstrating #China’s capability to train #largescalemodels without #Nvidia hardware. While LongCat-2.0 outperforms some Western models on specific benchmarks, it falls short on others. the-decoder.com/meituans-longc #AIagent #AI #ML #NLP #LLM #GenAI

  8. #Meituan trained a 1.6 trillion parameter AI model, #LongCat 2.0, entirely on #Chinesechips, demonstrating #China’s capability to train #largescalemodels without #Nvidia hardware. While LongCat-2.0 outperforms some Western models on specific benchmarks, it falls short on others. the-decoder.com/meituans-longc #AIagent #AI #ML #NLP #LLM #GenAI

  9. #Meituan trained a 1.6 trillion parameter AI model, #LongCat 2.0, entirely on #Chinesechips, demonstrating #China’s capability to train #largescalemodels without #Nvidia hardware. While LongCat-2.0 outperforms some Western models on specific benchmarks, it falls short on others. the-decoder.com/meituans-longc #AIagent #AI #ML #NLP #LLM #GenAI

  10. #Meituan trained a 1.6 trillion parameter AI model, #LongCat 2.0, entirely on #Chinesechips, demonstrating #China’s capability to train #largescalemodels without #Nvidia hardware. While LongCat-2.0 outperforms some Western models on specific benchmarks, it falls short on others. the-decoder.com/meituans-longc #AIagent #AI #ML #NLP #LLM #GenAI