B300 FP8: M3 EAGLE vs Off Speculative Decoding
Speculative decoding comparison of M3 EAGLE versus Off on B300 FP8 (NVIDIA Blackwell) running MiniMax M3 428B. Throughput, cost, and interactivity differences across LLM workloads. Use the chart controls below to switch sequences and metrics — same interactions as the main inference chart.
MTP acceptance-rate implementations differ across inference engines. Points from different engines are not directly comparable on the same curve — throughput and cost at matched interactivity may reflect engine-level differences rather than pure speculative decoding gains. Interpret cross-engine comparisons with caution.
M3 EAGLE posts 3329 tok/s/GPU for $0.20 per million tokens at 63 tok/s/user on MiniMax M3 428B (B300 FP8); Off posts 1299 tok/s/GPU for $0.50. M3 EAGLE is 156% cheaper per token; M3 EAGLE delivers 156% more tok/s/GPU. Draft-token acceptance rates determine whether speculative decoding helps or hurts at a given concurrency level.
Throughput at 105 tok/s/user on MiniMax M3 428B (B300 FP8): M3 EAGLE hits 1188 tok/s/GPU, Off hits 565. Per-million costs land at $0.55 and $1.15 respectively. M3 EAGLE is 110% cheaper per token; M3 EAGLE delivers 110% more tok/s/GPU. Speculative decoding trades extra compute on draft tokens for fewer decoding steps — the payoff depends on sequence length and batch size.
Toward the upper edge of the 22–189 tok/s/user interactivity band, at 148 tok/s/user on MiniMax M3 428B (B300 FP8): M3 EAGLE runs 1065 tok/s/GPU at $0.61/M tokens, Off runs 225 at $2.80/M. M3 EAGLE is 359% cheaper per token; M3 EAGLE delivers 373% more tok/s/GPU. Gains from speculative decoding vary by workload; short-output prompts tend to benefit less. (Numbers reflect this URL's pinned 1k/1k · fp8 workload — changing sequence or model updates both the table and chart; the table stays pinned to this page's precision, so precision toggles in the controls affect the chart only.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | M3 EAGLE:3328.8Off:1299.2 | M3 EAGLE:1187.8Off:565.1 | M3 EAGLE:1065.5Off:225.2 |
| Cost ($/M tok) | M3 EAGLE:$0.197Off:$0.505 | M3 EAGLE:$0.547Off:$1.151 | M3 EAGLE:$0.610Off:$2.800 |
| tok/s/MW | M3 EAGLE:1533993Off:598689 | M3 EAGLE:547386Off:260428 | M3 EAGLE:491004Off:103767 |
| Concurrency | M3 EAGLE:~117Off:~85 | M3 EAGLE:~34Off:~21 | M3 EAGLE:~16Off:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.