MiniMax just halved its token prices. Your AI agent bill may not follow.
MiniMax announced M3.1 Flash with halved per-token pricing and a 70 percent cost reduction claim for AI agents, plus inference nearly three times faster than M3. The company's figures come from different workloads than independent benchmarks, which show a 57 percent reduction on typical tasks. Real cost savings will depend on whether agents complete tasks in fewer loops, not just on lower per-token rates.
Key points
- MiniMax halved per-token prices for M3.1 Flash
- Claims 70% agent cost reduction, but varies by workload
- Independent benchmarks show 57% reduction on typical tasks
- Inference speed improved roughly 3x over M3
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. The New Stack did the reporting; read the original for the full story.
Read the full story at The New Stack
