MiniMax M3 428B — B200 vs B300
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and B300 (NVIDIA Blackwell) on MiniMax M3 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 →
B200 posts 36723 tok/s/chip for $0.01 per million tokens at 150 tok/s/user on MiniMax M3 428B; B300 posts 43400 tok/s/chip for $0.01. B200 is 11% cheaper per token; B300 delivers 18% more tok/s/chip.
Throughput at 233 tok/s/user on MiniMax M3 428B: B200 hits 19231 tok/s/chip, B300 hits 20592. Per-million costs land at $0.02 and $0.03 respectively. B200 is 22% cheaper per token; B300 delivers 7% more tok/s/chip.
B200 / B300 on MiniMax M3 428B at 316 tok/s/user: 7124 / 8014 tok/s/chip, $0.07 / $0.08 per million tokens. B200 is 16% cheaper per token; B300 delivers 12% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:36722.6B300:43399.5 | B200:19230.9B300:20591.7 | B200:7124.2B300:8013.8 |
| Cost ($/M tok) | B200:$0.013B300:$0.014 | B200:$0.025B300:$0.030 | B200:$0.067B300:$0.078 |
| tok/s/MW | B200:21475187B300:22841857 | B200:11246135B300:10837716 | B200:4166185B300:4217767 |
| Concurrency | B200:~22B300:~26 | B200:~12B300:~13 | B200:~3B300:~4 |
Total Tokens per $1 USD (Owning - Neocloud Giant) vs. Interactivity
MiniMax M3 428B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™
MiniMax M3 428B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™TCO $/chip/hr: VR200: 3.61H100: 1.55H200: 1.59B200: 2.07B300: 2.52GB200: 2.26GB300: 2.79MI300X: 1.16MI325X: 1.32MI355X: 2.09RTX6000PRO: 0.75Jalapeño (Teacup): 1.56
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
No data available
Please change the model, sequence, precision, date range or chip selection.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip
Interactivity is the rate at which a single user receives generated tokens while the model streams its answer — how quickly new words appear on screen. Higher values feel snappier; operators trade it against batch throughput.
How many total tokens (input + output) one US dollar of infrastructure spend buys, priced with the all-in hourly ownership cost of a Neocloud Giant operator. It is the reciprocal of cost per token, so higher means cheaper.
Formula: tok/$ = (total tok/s/chip × 3,600) ÷ all-in cost per chip-hour ($)