MiniMax 发布了其 M3 模型，这是一款拥有百万 token 上下文窗口和原生多模态能力的前沿级 AI。该模型在编码基准测试中表现强劲，在 SWE-Bench Pro 上达到 59.0%，在 Terminal Bench 2 上达到 66.0%。早期用户注意到其令人印象深刻的前端能力，以及在体验上与 Anthropic 的 Claude 相似，但性能和价值更优。 AI
影响 在编码基准测试中设定了新的 SOTA（State-of-the-Art），并为现有前沿模型提供了有竞争力的替代方案。
排序理由 前沿实验室模型发布，附带系统卡。
AI 生成摘要 · Google Gemini · 来自 23 个来源。 我们如何撰写摘要 →
RT @ZainHasan6: banger writeup on what it takes to make MiniMax M3 go brrrr! https://t.co/HJyaP8mT2N
Excited to bring M3 to more developers through @dgrid_ai 🤝 Frontier coding, native multimodality, and 1M-token context — now available on DGrid.
RT @arena: MiniMax M3 has landed in the Arena and has moved the Pareto frontier! Their latest model ranks #7 for Code Arena: Frontend, sco…
Mem0 is an official launch partner for MiniMax M3! M3's 1M token context window + @mem0ai 's memory layer = AI apps that truly remember. Build personalized AI agents with persistent memory, now with 50% off M3 during launch week. Get started with Minimax → https://t.co/KfgrtXg…
RT @RyanLeeMiniMax: 🔥 MiniMax-M3 just got faster. Thanks for all the excitement around MiniMax-M3 — the response has been far beyond our e…
MiniMax M3 now on GMI Cloud @gmi_cloud
MiniMax-M3 #6 overall on @ValsAI the new open-weight SOTA 🚀
Amazing deep dive from the @togethercompute team on serving MiniMax M3 in production. M3 with its 1M context, native multimodality and MiniMax Sparse Attention requires real work across paged decode, index scoring, and multimodal preprocessing to get it efficient. This is
RT @EasyRouterIO: MiniMax M3 is now available on and it’s 50% off this week.@MiniMax_AI 1M context · SOTA coding (SWE-Bench Pro 59%) · Na…
RT @CommandCodeAI: MiniMax M3 is live now on CommandCode! A frontier-class open-weight model with 1M context, frontier coding, agentic pe…
RT @kilocode: MiniMax M3 is live, and the benchmarks are turning heads. MiniMax reports 59.0% on SWE-Bench Pro, 66.0% on Terminal Bench 2.…
RT @flowith: flowith now supports minimax m3. 1m context. incredible coding ability. native multimodality. run and compare side-by-side.…
RT @happycapyai: MiniMax M3 @MiniMax_AI is now live on Happycapy 🎉 A major upgrade for agent workflows, especially when the task is messy,…
RT @lostinlatencyX: MiniMax M3’s frontend capabilities are pretty nice very strong model for the price. not lazy, thinks through the task…
RT @OnlyTerp: MiniMax-M3 in ultracode is INSANE 🤯 https://t.co/rAZuoJ7uXi
RT @JinjingLiang: Blown away by MiniMax M3 I’ve been using it for free inside @orca_build with the @opencode agent Mostly UI tasks & code…
RT @opencode: MiniMax M3 will be launching soon You can try it right now in OpenCode For free
MiniMax released its M3 flagship model, claiming it as the first domestic AI model to combine frontier coding, agentic capabilities, 1M-token context windows, and native multimodal processing in a single architecture.
<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://pub.towardsai.net/minimax-m3-decodes-1m-tokens-15x-faster-and-it-shouldnt-be-this-cheap-5428f2476957?source=rss----98111c9905da---4"><img src="https://cdn-images-1.medium.com/max/1672/1*_UF6l5PnzBPDKnsXGV…
<!-- SC_OFF --><div class="md"><p>Wow, not gonna lie, it feels like a clone of Claude, like, i feel like i'm really talking to Claude, but it's soooooo much better than M2.7. The MiniMax plan is kind of expensive now and usage spiked up a lot more than M2.7 did, but i think def w…
MiniMax released M3 with a 1M-token context window and native multimodal support via API, but the open-weight claim remains conditional. Model files and technical documentation arrive in 10 days. Developers get API access first; local inspection comes later. https://www. implicat…
<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1ttdkzu/minimax_m3_has_been_released/"> <img alt="Minimax M3 has been released" src="https://external-preview.redd.it/Ym4sCzj2wR3hrL4cqGefONXtmu3aFvVWJ1IS-DAt1sk.png?width=140&height=78&auto=webp&…
<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1ttc5u3/minimax_m3_is_starting_to_rollout_on_the_api/"> <img alt="MiniMax M3 is starting to rollout on the API" src="https://preview.redd.it/goe30iwkek4h1.png?width=140&height=27&auto=webp&s=a143e…