Expired50% confidenceFactExact time
STEP3.7 Flash features end-to-end multimodal real-time interaction capabilities and can simultaneously process multiple input modalities including text, images, audio, and video.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
StepFun STEP3.7 Flash Tops AA Benchmark — Multimodal Reasoning Speed Takes Off
bilibiliAI天天酱6/6/2026
Related Claims
UnverifiedThe Flash version focuses on real-time interaction, is suitable for latency-sensitive scenarios, and supports streaming output.76% similarUnverifiedStep 3.5 Flash采用混合注意力系统（Hybrid Attention），对短内容使用滑动窗口注意力，对长内容切换为完整注意力机制75% similarUnverifiedStepFun AI 正式发布了 Step 3.7 Flash，这是一款拥有 198B 参数的稀疏混合专家（Sparse MoE）视觉语言模型72% similarUnverifiedThe Flash model is positioned as lightweight, fast, and low-cost, suitable for scenarios such as high-concurrency requests, content classification, summary generation, customer service Q&A, and batch data processing.68% similarUnverified盘古2.0 Flash提供512K的超长上下文窗口，采用MLA、DSA和SWA三种注意力机制的组合设计66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/53709API
curl https://kongchang.com/api/v1/knowledge/claims/53709MCP
get_claim(id=53709)