InferenceBench

Qwen/Qwen2-VL-7B-Instruct

vllm unknown on 8x NVIDIA H100 80GB HBM3 (fp16) · seed 0 · run 019e3b85-73f3-7c52-81c3-0e2b87c0acc0 · signed

Headline metrics

TTFT P50 (ms) 160
TTFT P99 (ms)
Throughput (tok/s)
$/M tokens
J/token
Power avg (W)
Power peak (W)
WER mean
J / audio s

All metrics

accuracy 0.8000
accuracy_p05 0.2000
accuracy_p50 1.00
accuracy_p95 1.00
n_ok 5.00
n_samples 5.00
ok_rate 1.00
tokens_out_total 36.00
total_p50_ms 160
ttft_p50_ms 160

Provenance

Model revisionunknown00
Providervllm
Hardware fingerprint550474fc9132129654f5d20c316eaffec99a1f67c0cc0b4641e60efb46ea79e7
Datasetbuiltin-ocr-mini · 7b1c3328af108dd2aada94d9d3c0c2c7f71b97f740627d481f9c926c2253c160
Timestamp2026-05-18T14:37:42.259556+00:00
Envelope JSONca49934392e9.json

Verify this result

bench verify /bench/leaderboard/envelopes/ca49934392e9.json

bench verify re-downloads this envelope, recomputes the canonical content hash, and validates the Sigstore signature against the embedded certificate + Rekor entry.