Round score
0.6316
+63.16%
Prompts scored
26 / 32
median_e2e_speedup
Below tolerance
6
answered too little to score
Zeroed total
6
scored 0.0 without measuring
The round score is the scoring rule applied to the prompts below. Both engines are compared at the same output token count, so a candidate that stops early cannot buy speed by answering less: a prompt under the tolerance bar scores nothing and shows its gate reason.
32 prompts
| Prompt | Speedup | Baseline | Candidate | Tokens | Gate |
|---|---|---|---|---|---|
| hf-000 | +67.84% | 0.661s | 0.213s | 42 | — |
| hf-001 | +70.56% | 0.722s | 0.213s | 43 | — |
| hf-002 | +70.61% | 0.727s | 0.214s |
32 requests
| Request | TTFT | Total | Tokens | Slowest gap | Stream |
|---|---|---|---|---|---|
| hf-000 | 74.39ms | 0.213s | 42 | 15.83ms | — |
| hf-001 | 88.94ms | 0.213s | 43 | 16.18ms | — |
| hf-002 | 75.70ms | 0.253s |
verdict
evidence
min logprob
num prompts
mean logprob
coverage ratio
quantile logprob
num positions scored
role
| 47 |
| — |
| hf-003 | +63.51% | 0.691s | 0.252s | 47 | — |
| hf-004 | +52.04% | 0.589s | 0.282s | 40 | — |
| hf-005 | +65.00% | 0.811s | 0.284s | 54 | — |
| hf-006 | +70.86% | 0.713s | 0.208s | 41 | — |
| hf-008 | +62.56% | 0.512s | 0.192s | 26 | — |
| hf-007 | +66.49% | 0.895s | 0.300s | 52 | — |
| hf-009 | +64.18% | 0.714s | 0.256s | 42 | — |
| hf-010 | +66.54% | 0.763s | 0.255s | 47 | — |
| hf-011 | — | — | — | 42 | candidate output below tolerance |
| hf-012 | +53.15% | 0.750s | 0.351s | 42 | — |
| hf-013 | — | — | — | 42 | candidate output below tolerance |
| hf-014 | +70.68% | 0.699s | 0.205s | 40 | — |
| hf-015 | +71.86% | 1.101s | 0.310s | 63 | — |
| hf-016 | +68.91% | 0.911s | 0.283s | 52 | — |
| hf-017 | +71.91% | 0.769s | 0.216s | 44 | — |
| hf-018 | — | — | — | 27 | candidate output below tolerance |
| hf-019 | +66.90% | 0.754s | 0.250s | 42 | — |
| hf-021 | — | — | — | 23 | candidate output below tolerance |
| hf-020 | +70.00% | 1.112s | 0.334s | 61 | — |
| hf-022 | — | — | — | 41 | candidate output below tolerance |
| hf-023 | +62.82% | 0.757s | 0.282s | 41 | — |
| hf-025 | +59.68% | 0.753s | 0.304s | 43 | — |
| hf-024 | +58.83% | 0.756s | 0.311s | 41 | — |
| hf-026 | — | — | — | 17 | candidate output below tolerance |
| hf-028 | +46.48% | 0.527s | 0.282s | 35 | — |
| hf-027 | +73.93% | 0.798s | 0.208s | 53 | — |
| hf-029 | +58.15% | 0.591s | 0.247s | 44 | — |
| hf-031 | +59.12% | 0.524s | 0.214s | 43 | — |
| hf-030 | +50.69% | 0.809s | 0.399s | 68 | — |
| 48 |
| 39.34ms |
| — |
| hf-003 | 91.25ms | 0.252s | 47 | 38.18ms | — |
| hf-004 | 82.08ms | 0.282s | 40 | 39.28ms | — |
| hf-005 | 82.38ms | 0.317s | 69 | 39.53ms | — |
| hf-006 | 83.72ms | 0.285s | 51 | 39.41ms | — |
| hf-007 | 83.63ms | 0.300s | 52 | 41.00ms | — |
| hf-008 | 83.71ms | 0.318s | 52 | 45.27ms | — |
| hf-009 | 85.85ms | 0.290s | 47 | 46.42ms | — |
| hf-010 | 89.85ms | 0.255s | 47 | 42.49ms | — |
| hf-011 | 89.61ms | 0.213s | 42 | 16.30ms | — |
| hf-012 | 146.03ms | 0.351s | 42 | 69.58ms | — |
| hf-013 | 119.41ms | 0.282s | 42 | 38.38ms | — |
| hf-014 | 82.26ms | 0.205s | 43 | 15.87ms | — |
| hf-015 | 80.17ms | 0.310s | 63 | 70.53ms | — |
| hf-016 | 100.98ms | 0.283s | 52 | 43.84ms | — |
| hf-017 | 92.61ms | 0.216s | 44 | 16.01ms | — |
| hf-018 | 88.41ms | 0.180s | 27 | 15.53ms | — |
| hf-019 | 86.51ms | 0.250s | 42 | 39.51ms | — |
| hf-020 | 85.49ms | 0.479s | 93 | 68.35ms | — |
| hf-021 | 87.55ms | 0.167s | 23 | 16.25ms | — |
| hf-022 | 112.71ms | 0.295s | 41 | 67.17ms | — |
| hf-023 | 107.24ms | 0.282s | 41 | 66.73ms | — |
| hf-024 | 120.75ms | 0.311s | 41 | 67.70ms | — |
| hf-025 | 111.03ms | 0.304s | 43 | 38.18ms | — |
| hf-026 | 72.88ms | 0.120s | 17 | 16.23ms | — |
| hf-027 | 69.81ms | 0.208s | 53 | 15.96ms | — |
| hf-028 | 124.82ms | 0.361s | 58 | 39.66ms | — |
| hf-029 | 85.09ms | 0.247s | 46 | 38.22ms | — |
| hf-030 | 87.13ms | 0.460s | 89 | 40.54ms | — |
| hf-031 | 88.10ms | 0.214s | 43 | 16.44ms | — |
evidence
requests per s
sla goodput ratio
output tokens per s
p99 e2e ms rel range
p99 itl ms rel range
p99 ttft ms rel range
Image digest
Patch hash