SGLang speeds up MiniMax-H3 1.95x with zero quality loss
LMSYS benchmarked MiniMax-H3 on 8x H200 GPUs: SGLang's lossless path beats Diffusers by up to 1.95x, while steeper tiers trade SSIM fidelity for up to 6x speed.
MiniMax News and Analysis
10 verified stories covering MiniMax, product updates and industry developments.
MiniMax launches Design, stitching 15-second clips into full films
MiniMax's new Design platform uses AI agents to stitch its 15-second H3 clips into coherent films up to 3 minutes, aimed at short dramas and brand video.
MiniMax Opens Its H3 Video Model
MiniMax H3 combines 2K, 15-second video, native stereo audio, RMB 0.8-per-second pricing and a coming open-weight release.
MiniMax raises another HK$16B
MiniMax raises another HK$16B. A concise localization of the verified facts and the industry signal behind the story.
Zhipu and MiniMax Face Hong Kong Lock-Up Test
Zhipu and MiniMax shares fell before cornerstone investor lock-ups expire on July 8 and 9, raising concerns about a sudden supply of tradable stock.
MiniMax M3 brings sparse attention, supports million-token context
MiniMax released the open-weight M3 model with 1M-token context, advanced coding ability, and native multimodality, claiming it is the first to combine all three.
MiniMax M2.7 Self-Iterates 100 Rounds, Performance Up 30%
MiniMax's M2.7 model used an agent framework to automatically run over 100 training iterations without human intervention, boosting internal benchmarks by 30%. The 230B-parameter sparse MoE model activates only 10B per inference and is now open-source.
MiniMax Releases M1 Reasoning Model
MiniMax has unveiled its M1 reasoning model series, targeting OpenAI's o-series and DeepSeek's R1 with strong performance in math, code, and logic tasks, while its M2.5 iteration scores 80.2% on SWE-bench Verified.
MiniMax's Hailuo AI Emerges as a Dark Horse in Chinese Video Generation
MiniMax, a relatively low-profile Chinese AI company, has built a strong reputation for its Hailuo AI video generation tool, which rivals top-tier domestic tools and early versions of Sora in quality.
MiniMax-01 Uses Lightning Attention to Reduce Compute Power
MiniMax's 01 series introduces Lightning Attention, a linear attention variant that cuts computational complexity from O(n²) to near O(n), delivering significant speed gains in long-context scenarios while maintaining competitive accuracy.