MiniMax M3.1 Flash shows China’s AI price race has no floor in sight
What happened
MiniMax released the M3.1 Flash, a new AI model that scores 42 on the Artificial Analysis Intelligence Index, 13 points higher than its predecessor. It costs half as much while maintaining strong performance. On the DeepSWE v1.1 coding benchmark, the M3.1 Flash achieves 73.7 percent accuracy at under $0.30 per task. This matches the performance and price of the Deepseek V4.1 Flash 1017 model, revealing a steep descent in AI model pricing coming out of China.
Why it matters
AI pricing is under intense pressure as MiniMax pushes model cost down sharply while improving accuracy. For businesses and developers, this means powerful AI is becoming accessible at much lower prices without sacrificing quality. The comparison to Deepseek shows China’s AI race is tightening the gap on performance-cost efficiency globally. Operators buying AI compute and services should expect more aggressive pricing and accelerated innovation as this competition cuts into traditional market pricing floors, forcing suppliers elsewhere to re-examine cost structures.
What to watch next
Keep an eye on how fast MiniMax and other Chinese AI companies continue driving prices down. This will test the durability of AI pricing in the West and among global cloud providers. The M3.1 Flash performance sets a new baseline for cost-effective coding tasks, so competitors will need to respond or risk losing share among cost-conscious developers and businesses. Investors and buyers should watch emerging AI benchmarks for shifts in value delivered per dollar and how this impacts AI adoption economics across sectors.
AI Quick Briefs Editorial Desk