Round score
0.6841
+68.41%
Prompts scored
27 / 32
median_e2e_speedup
Below tolerance
5
answered too little to score
Zeroed total
5
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 | +57.66% | 0.736s | 0.311s | 48 | — |
| hf-001 | +69.77% | 0.723s | 0.219s | 42 | — |
| hf-002 | +69.80% | 0.722s | 0.218s |
32 requests
| Request | TTFT | Total | Tokens | Slowest gap | Stream |
|---|---|---|---|---|---|
| hf-000 | 112.66ms | 0.311s | 48 | 66.76ms | — |
| hf-001 | 95.78ms | 0.257s | 52 | 38.78ms | — |
| hf-002 | 95.71ms | 0.218s |
verdict
evidence
min logprob
num prompts
mean logprob
coverage ratio
quantile logprob
num positions scored
role
| 42 |
| — |
| hf-003 | +62.92% | 0.915s | 0.339s | 48 | — |
| hf-004 | +55.77% | 0.782s | 0.346s | 43 | — |
| hf-005 | +68.56% | 0.785s | 0.247s | 42 | — |
| hf-006 | — | — | — | 27 | candidate output below tolerance |
| hf-007 | +77.59% | 0.783s | 0.176s | 42 | — |
| hf-008 | +60.28% | 0.784s | 0.311s | 42 | — |
| hf-009 | +73.65% | 0.775s | 0.204s | 43 | — |
| hf-011 | — | — | — | 42 | candidate output below tolerance |
| hf-010 | +72.87% | 0.738s | 0.200s | 42 | — |
| hf-012 | +72.68% | 0.763s | 0.208s | 43 | — |
| hf-015 | +56.14% | 0.342s | 0.150s | 16 | — |
| hf-013 | — | — | — | 28 | candidate output below tolerance |
| hf-014 | +79.06% | 0.901s | 0.189s | 46 | — |
| hf-016 | +75.19% | 0.768s | 0.191s | 43 | — |
| hf-018 | +74.35% | 0.741s | 0.190s | 40 | — |
| hf-017 | — | — | — | 43 | candidate output below tolerance |
| hf-021 | +72.85% | 0.760s | 0.206s | 40 | — |
| hf-019 | +76.01% | 1.326s | 0.318s | 73 | — |
| hf-020 | — | — | — | 23 | candidate output below tolerance |
| hf-022 | +72.60% | 0.780s | 0.214s | 42 | — |
| hf-023 | +61.54% | 0.740s | 0.285s | 39 | — |
| hf-024 | +68.26% | 0.810s | 0.257s | 45 | — |
| hf-025 | +74.17% | 0.780s | 0.201s | 43 | — |
| hf-027 | +71.58% | 0.538s | 0.153s | 28 | — |
| hf-026 | +74.46% | 0.771s | 0.197s | 42 | — |
| hf-029 | +54.02% | 0.710s | 0.326s | 43 | — |
| hf-028 | +64.83% | 0.946s | 0.333s | 59 | — |
| hf-030 | +51.45% | 0.615s | 0.299s | 41 | — |
| hf-031 | +59.78% | 0.550s | 0.221s | 43 | — |
| 45 |
| 16.15ms |
| — |
| hf-003 | 109.76ms | 0.339s | 48 | 75.51ms | — |
| hf-004 | 139.10ms | 0.346s | 44 | 72.99ms | — |
| hf-005 | 115.54ms | 0.247s | 42 | 38.18ms | — |
| hf-006 | 86.52ms | 0.179s | 27 | 16.09ms | — |
| hf-007 | 67.67ms | 0.176s | 42 | 16.15ms | — |
| hf-008 | 137.10ms | 0.311s | 42 | 43.85ms | — |
| hf-009 | 95.62ms | 0.204s | 43 | 16.09ms | — |
| hf-010 | 77.42ms | 0.200s | 42 | 15.78ms | — |
| hf-011 | 117.47ms | 0.286s | 42 | 41.37ms | — |
| hf-012 | 85.96ms | 0.208s | 43 | 16.11ms | — |
| hf-013 | 82.39ms | 0.190s | 28 | 16.04ms | — |
| hf-014 | 65.67ms | 0.189s | 46 | 16.36ms | — |
| hf-015 | 88.63ms | 0.166s | 20 | 15.99ms | — |
| hf-016 | 67.89ms | 0.191s | 43 | 16.15ms | — |
| hf-017 | 66.16ms | 0.189s | 43 | 16.08ms | — |
| hf-018 | 67.33ms | 0.190s | 42 | 16.01ms | — |
| hf-019 | 56.22ms | 0.318s | 73 | 42.69ms | — |
| hf-020 | 86.72ms | 0.164s | 23 | 16.14ms | — |
| hf-021 | 98.23ms | 0.206s | 40 | 15.89ms | — |
| hf-022 | 90.79ms | 0.214s | 42 | 16.69ms | — |
| hf-023 | 98.33ms | 0.285s | 39 | 78.03ms | — |
| hf-024 | 112.76ms | 0.257s | 45 | 40.45ms | — |
| hf-025 | 78.19ms | 0.201s | 43 | 16.21ms | — |
| hf-026 | 88.01ms | 0.197s | 42 | 16.38ms | — |
| hf-027 | 75.10ms | 0.199s | 44 | 16.93ms | — |
| hf-028 | 78.38ms | 0.349s | 61 | 70.69ms | — |
| hf-029 | 117.39ms | 0.326s | 43 | 72.74ms | — |
| hf-030 | 126.47ms | 0.299s | 41 | 39.90ms | — |
| hf-031 | 82.40ms | 0.221s | 43 | 16.54ms | — |
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