模型对比
M3 和 Step 3.7 Flash 都是 AI 领域的模型。以下是基于 traeai 收录的真实报道数据的全面对比。
模型
也叫：Minimax M3
MiniMax 推出的新一代大语言模型，用于编码等任务。
4 篇相关报道
模型
也叫：step3.7flash
阶跃星辰发布的高效推理模型。
7 篇相关报道
4
M3 相关
0
共同提及
7
Step 3.7 Flash 相关
Minimax M3 combines 1M context, multimodality, open weights, and agentic coding — currently free via Open Code; scored 4/10 on complex UI tasks, ideal for toolchain validation but not high-fidelity state systems.
入选理由：M3 支持 100 万 token 上下文窗口，采用稀疏注意力架构控制成本，专为编码代理设计。
The M3 model by MiniMax is now available on Ollama Cloud, deployed in the US with zero data retention, optimized for coding and agentic tasks. It achieves 59.0%+ on SWE-Bench Pro and supports up to 1M context length via sparse attention.
入选理由：M3 在 SWE-Bench Pro 基准中取得 59.0% 正确率，优于多数开源模型。
MiniMax M3 is now available on Vercel AI Gateway, offering a 1M-token context window and native multimodal support via MSA architecture, enhancing engineering efficiency and tool integration.
入选理由：M3模型支持1M令牌上下文窗口，显著提升长文本处理能力。
MiniMax M3 is about to launch; a developer will conduct deep testing using their own coding agent and harness, with a review to follow. Free trial available on OpenCode.
入选理由：MiniMax M3 即将发布，预计在 OpenCode 平台上线。
StepFun released Step 3.7 Flash — a high-efficiency agentic coding model supporting multimodal understanding, tool use, and long-running workflows; its standout feature is full free access in Hermes Agent, removing typical API/credit barriers for real-world testing.
入选理由：Step 3.7 Flash 是 StepFun 新一代 agentic coding 模型，含196B总参数 + 1.8B 视觉模块 + ~11B 激活参数，支持256K上下文窗口。
Step 3.7 Flash by Yujue Star is a new-generation Flash model for production-grade AI Agents, featuring native multimodal understanding, high throughput with low latency, and enhanced web search. It achieves 97% of Claude Opus 4.6's coding performance at only 1/9 the cost per task, ideal for high-frequency, complex real-world workflows.
入选理由：Step 3.7 Flash 采用稀疏 MoE 架构，激活参数仅 11B，最高生成速度达 400 Tokens/s，支持 40 个 Agent 并行运行。
Step 3.7 Flash is a 196B MoE model designed from the ground up for inference efficiency, using MFA and AFD techniques to reduce KV-cache usage to ~22% of DeepSeek, supporting agent, coding, and multimodal workflows, open-sourced under Apache 2.0 and available on Fireworks.
入选理由：Step 3.7 Flash 是 196B MoE 模型，从设计之初就聚焦推理效率，而非事后优化。
NVIDIA introduces Step 3.7 Flash for large-scale MoE models.
入选理由：Step 3.7 Flash 支持 198B 参数 MoE 模型，其中 11B 参数为活跃状态。
文章仅提供了一个链接，介绍如何在 NVIDIA GPU 上运行 Step 3.7 Flash，缺乏深度和技术细节。
入选理由：文章链接指向 NVIDIA 开发者博客，介绍 Step 3.7 Flash 的运行方法。
Step 3.7 Flash significantly reduces KV-cache cost via MFA + AFD technology, enabling efficient inference with one-click deployment.
入选理由：Step 3.7 Flash采用MFA + AFD技术，将KV-cache成本降至原模型的分数。
Step 3.7 Flash model achieves 400 tokens/sec processing speed, reducing Agent task cost to a fraction of Claude's.
入选理由：处理速度达400 tokens/秒
