MiniMax M3 428B · Speculative decoding
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 3,329 tok/s/chip for $0.19 per million tokens at 63 tok/s/user on MiniMax M3 428B (B300 FP8); Off posts 1,299 tok/s/chip for $0.48. M3 EAGLE delivers 156% more tok/s/chip on the same chip, which makes it 61% cheaper per token. 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 1,188 tok/s/chip, Off hits 565. Costs per million tokens land at $0.53 and $1.11 respectively. M3 EAGLE delivers 110% more tok/s/chip on the same chip, which makes it 52% cheaper per token. 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 1,065 tok/s/chip at $0.59 per million tokens, Off runs 225 at $2.79. M3 EAGLE delivers 4.7× the tok/s/chip on the same chip, which makes it 79% cheaper per token. Gains from speculative decoding vary by workload; short-output prompts tend to benefit less. (Numbers reflect this URL's pinned 1k/1k · fp8 workload — changing sequence or model updates both the table and chart; the table stays pinned to this page's precision, so precision toggles in the controls affect the chart only.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | M3 EAGLE3,329Off1,299 | M3 EAGLE1,188Off565 | M3 EAGLE1,065Off225 |
| Cost ($/M tok) | M3 EAGLE$0.19Off$0.48 | M3 EAGLE$0.53Off$1.11 | M3 EAGLE$0.59Off$2.79 |
| tok/s/MW | M3 EAGLE1,751,981Off683,765 | M3 EAGLE625,173Off297,436 | M3 EAGLE560,778Off118,512 |
| Concurrency | M3 EAGLE~117Off~85 | M3 EAGLE~34Off~21 | M3 EAGLE~16Off~7 |
