@sdrzn: MiniMax's new m3 model scores the same as opus 4.7 on terminal-bench 2.1 at 1/20th the compute/cost of their previous m…
Summary
MiniMax's new m3 model achieves the same score as Opus 4.7 on terminal-bench 2.1 while using 1/20th the compute and cost, attributed to their novel MiniMax Sparse Attention architecture.
View Cached Full Text
Cached at: 06/01/26, 11:19 AM
MiniMax’s new m3 model scores the same as opus 4.7 on terminal-bench 2.1 at 1/20th the compute/cost of their previous model.
Their blog credits this to their new ‘MiniMax Sparse Attention’ architecture next to this crazy diagram.
Here’s a breakdown of how this works 🧵 https://t.co/eZuNQ5Ju67
Similar Articles
@PrajwalTomar_: Everyone's sleeping on MiniMax. Again. They just shipped M3. The first open-weights model to combine frontier coding, 1…
MiniMax released M3, an open-weights model combining frontier coding, 1M context, and native multimodality, offering comparable performance to Opus at a fraction of the cost.
MiniMax teases upcoming M3 model with new sparse attention mechanism and 15.6X long-context response speed boost (12 minute read)
MiniMax has released a detailed technical report on its M2 series and teased the upcoming M3 model, which uses a novel sparse attention mechanism to achieve up to 15.6× faster decoding at million-token contexts.
@jiayuan_jy: A few objective clarifications: 1) This post has nothing to do with MiniMax (I never take sponsored posts). 2) 'Subjective feel' is not the same as actual performance; it's not quantitative data. After more extensive experience, overall coding ability is a qualitative improvement compared to m2.7. A current shortcoming is that 1-shot results compared with...
Jiayuan Zhang shared his initial experience with the M3 model's coding ability, stating that it is a qualitative improvement compared to m2.7, but the 1-shot results are not as comprehensive as Opus 4.6/4.7 and GPT5.5.
@TeksEdge: With MiniMax M3 open source now out, here is what to expect on quants and sizes, including VRAM needed: MiniMax M3 (428…
MiniMax M3, a 428B MoE model with ~23B active parameters, is now open source. It offers ultra-long context (up to 1M) and efficiency improvements, with various quantized sizes and VRAM requirements for local deployment.
MiniMax M3 (2 minute read)
MiniMax introduces M3, the first open-weights model to combine coding, agentic, and multimodal capabilities with up to 1M context via sparse attention.