InferenceBench

microsoft/Phi-3.5-mini-instruct

vllm unknown on 8x NVIDIA H100 80GB HBM3 (fp16) · seed 0 · run 019e3b5d-39b2-7763-a9c2-b05da554850d · signed

Headline metrics

TTFT P50 (ms) 11.23
TTFT P99 (ms)
Throughput (tok/s)
$/M tokens
J/token
Power avg (W)
Power peak (W)
WER mean
J / audio s

All metrics

accuracy 0.2000
accuracy_p05 0.0000
accuracy_p50 0.0000
accuracy_p95 1.00
n_ok 10.00
n_samples 10.00
ok_rate 1.00
tokens_out_total 286
total_p50_ms 192
ttft_p50_ms 11.23

Provenance

Model revisionunknown00
Providervllm
Hardware fingerprint550474fc9132129654f5d20c316eaffec99a1f67c0cc0b4641e60efb46ea79e7
Datasetbuiltin-reasoning-mini · 0968fee174ea0335f916f4bce969d72d59090d48ac46fb471901fc033be4eb1d
Timestamp2026-05-18T13:53:45.906434+00:00
Envelope JSON3db3db19d903.json

Verify this result

bench verify /bench/leaderboard/envelopes/3db3db19d903.json

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