MiniMax M3.1 Flash 428B · Chip comparison
MiniMax M3.1 Flash 428B — H200 vs Vera Rubin NVL72
Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on MiniMax M3.1 Flash 428B. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX →
At 308 tok/s/user interactivity on MiniMax M3.1 Flash 428B, H200 delivers 2,394 tok/s/chip at $0.14 per million tokens; Vera Rubin NVL72 delivers 254,869 tok/s/chip at $0.004. Vera Rubin NVL72 is 97% cheaper per token; Vera Rubin NVL72 delivers 106× the tok/s/chip at this point.
H200 posts 2,394 tok/s/chip for $0.14 per million tokens at 607 tok/s/user on MiniMax M3.1 Flash 428B; Vera Rubin NVL72 posts 61,932 tok/s/chip for $0.016. Vera Rubin NVL72 is 89% cheaper per token; Vera Rubin NVL72 delivers 26× the tok/s/chip.
Throughput at 906 tok/s/user on MiniMax M3.1 Flash 428B: H200 hits 2,394 tok/s/chip, Vera Rubin NVL72 hits 32,135. Costs per million tokens land at $0.14 and $0.031 respectively. Vera Rubin NVL72 is 78% cheaper per token; Vera Rubin NVL72 delivers 13× the tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H2002,394Vera Rubin NVL72254,869 | H2002,394Vera Rubin NVL7261,932 | H2002,394Vera Rubin NVL7232,135 |
| Cost ($/M tok) | H200$0.14Vera Rubin NVL72$0.004 | H200$0.14Vera Rubin NVL72$0.016 | H200$0.14Vera Rubin NVL72$0.031 |
| tok/s/MW | H2001,747,420Vera Rubin NVL7277,232,975 | H2001,747,420Vera Rubin NVL7218,767,203 | H2001,747,420Vera Rubin NVL729,737,835 |
| Concurrency | H200~1Vera Rubin NVL72~640 | H200~1Vera Rubin NVL72~63 | H200~1Vera Rubin NVL72~11 |