精选理由
MiniMax-M3 让智能体一口气带进长历史、图、视频、文档和工具输出,Together 优化后每 GPU token 翻倍,自动任务成本更低。
MiniMax-M3 模型支持智能体携带长历史(超过百万token)、图像、视频、文档和工具输出进入上下文,显著提升多模态信息处理能力。Together 的推理优化通过改进服务路径上的 token 吞吐量,使这一能力在大规模部署时更实用。相比之前方案,每 GPU 可处理更多 token,从而降低每美元自动化工作成本。
原文 · Together AI
MiniMax-M3 expands what agents can carry into cont…
MiniMax-M3 expands what agents can carry into context: long histories, images, video, documents, and tool outputs.
Together’s inference work makes that practical at scale by improving token throughput across the serving path.
More tokens per GPU means more work automated per dollar.
We go deeper in this deep dive from @ywangfirstlean, @zhyncs42, @realDanFu and the team.