Unverified50% confidenceFactExact time
StepFun is one of the domestic foundational large model R&D teams, with its Step series covering multiple directions including language and multimodality.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/19/2026
First Seen
Valid until: 12/18/2026
Sources
Related Entities
Related Claims
UnverifiedThe large language model Step 5 Preview from StepFun has reportedly entered the Pareto frontier of Artificial Analysis.71% similarUnverifiedGemini 3.5 Live Translate 采用端到端学习范式，将ASR、NLU、MT、TTS四个环节在同一多模态大模型中联合优化63% similarUnverifiedGemini架构通过视觉编码器（Vision Encoder）与语言模型的联合推理实现跨模态对齐（Cross-modal Alignment）能力61% similarUnverified视觉大模型发展的原生统一架构从预训练阶段就将视觉Token与文本Token置于同一序列空间联合建模，代表性工作包括Google的Gemini系列与Meta的Chameleon61% similarUnverifiedNotebookLM is underpinned by the Gemini series of large models for long-context understanding, supporting input windows at the scale of hundreds of thousands of tokens.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/934776API
curl https://kongchang.com/api/v1/knowledge/claims/934776MCP
get_claim(id=934776)