Round score
0.7251
+72.51%
Prompts scored
29 / 32
median_e2e_speedup
Below tolerance
3
answered too little to score
Zeroed total
3
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 | +72.57% | 0.713s | 0.196s | 42 | — |
| hf-001 | +73.53% | 0.732s | 0.194s | 42 | — |
| hf-002 | +77.34% | 0.737s | 0.167s |
32 requests
| Request | TTFT | Total | Tokens | Slowest gap | Stream |
|---|---|---|---|---|---|
| hf-000 | 84.02ms | 0.196s | 42 | 17.84ms | — |
| hf-001 | 82.30ms | 0.194s | 42 | 17.21ms | — |
| hf-002 | 71.50ms | 0.167s |
verdict
evidence
min logprob
num prompts
mean logprob
coverage ratio
quantile logprob
num positions scored
role
| 42 |
| — |
| hf-003 | +79.78% | 0.803s | 0.162s | 47 | — |
| hf-004 | — | — | — | 28 | candidate output below tolerance |
| hf-006 | +69.11% | 0.468s | 0.145s | 26 | — |
| hf-005 | — | — | — | 28 | candidate output below tolerance |
| hf-007 | +64.85% | 0.597s | 0.210s | 42 | — |
| hf-008 | +59.06% | 0.723s | 0.296s | 48 | — |
| hf-009 | +73.44% | 0.716s | 0.190s | 44 | — |
| hf-010 | +69.93% | 0.707s | 0.213s | 40 | — |
| hf-011 | +69.91% | 0.712s | 0.214s | 41 | — |
| hf-012 | +72.91% | 0.729s | 0.198s | 42 | — |
| hf-013 | +75.44% | 0.734s | 0.180s | 42 | — |
| hf-014 | +78.79% | 0.738s | 0.156s | 42 | — |
| hf-015 | +77.20% | 0.765s | 0.174s | 45 | — |
| hf-017 | — | — | — | 28 | candidate output below tolerance |
| hf-016 | +75.10% | 0.991s | 0.247s | 61 | — |
| hf-018 | +75.68% | 0.898s | 0.219s | 54 | — |
| hf-019 | +69.09% | 0.698s | 0.216s | 42 | — |
| hf-020 | +73.73% | 0.736s | 0.193s | 42 | — |
| hf-021 | +78.05% | 0.752s | 0.165s | 44 | — |
| hf-022 | +71.00% | 0.735s | 0.213s | 42 | — |
| hf-023 | +59.26% | 0.717s | 0.292s | 40 | — |
| hf-024 | +72.97% | 0.750s | 0.203s | 43 | — |
| hf-025 | +74.03% | 0.732s | 0.190s | 42 | — |
| hf-027 | +72.45% | 0.503s | 0.139s | 26 | — |
| hf-026 | +72.74% | 0.735s | 0.200s | 42 | — |
| hf-028 | +65.99% | 0.749s | 0.255s | 43 | — |
| hf-029 | +70.21% | 0.676s | 0.201s | 42 | — |
| hf-030 | +56.28% | 0.659s | 0.288s | 47 | — |
| hf-031 | +63.31% | 0.540s | 0.198s | 43 | — |
| 42 |
| 17.13ms |
| — |
| hf-003 | 66.80ms | 0.162s | 47 | 17.11ms | — |
| hf-004 | 67.36ms | 0.147s | 28 | 17.17ms | — |
| hf-005 | 66.99ms | 0.163s | 28 | 17.58ms | — |
| hf-006 | 63.76ms | 0.145s | 26 | 17.72ms | — |
| hf-007 | 82.01ms | 0.210s | 42 | 16.99ms | — |
| hf-008 | 126.93ms | 0.296s | 48 | 43.55ms | — |
| hf-009 | 96.45ms | 0.190s | 44 | 16.43ms | — |
| hf-010 | 86.36ms | 0.213s | 40 | 17.74ms | — |
| hf-011 | 102.77ms | 0.214s | 41 | 17.05ms | — |
| hf-012 | 86.31ms | 0.198s | 47 | 17.68ms | — |
| hf-013 | 82.74ms | 0.180s | 42 | 17.41ms | — |
| hf-014 | 76.09ms | 0.156s | 42 | 17.43ms | — |
| hf-015 | 63.96ms | 0.174s | 45 | 17.58ms | — |
| hf-016 | 62.88ms | 0.247s | 61 | 40.66ms | — |
| hf-017 | 86.17ms | 0.182s | 28 | 17.48ms | — |
| hf-018 | 75.81ms | 0.219s | 54 | 17.39ms | — |
| hf-019 | 89.16ms | 0.216s | 42 | 17.03ms | — |
| hf-020 | 97.38ms | 0.193s | 42 | 16.82ms | — |
| hf-021 | 69.00ms | 0.165s | 44 | 17.36ms | — |
| hf-022 | 86.37ms | 0.252s | 52 | 38.99ms | — |
| hf-023 | 84.93ms | 0.292s | 40 | 48.83ms | — |
| hf-024 | 92.07ms | 0.203s | 43 | 17.04ms | — |
| hf-025 | 94.39ms | 0.206s | 51 | 17.27ms | — |
| hf-026 | 88.62ms | 0.200s | 42 | 17.12ms | — |
| hf-027 | 74.67ms | 0.216s | 67 | 16.91ms | — |
| hf-028 | 88.13ms | 0.255s | 43 | 39.92ms | — |
| hf-029 | 90.69ms | 0.201s | 42 | 17.94ms | — |
| hf-030 | 113.97ms | 0.288s | 47 | 39.32ms | — |
| hf-031 | 87.88ms | 0.198s | 44 | 16.87ms | — |
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