InferenceBench

meta-llama/Llama-3.1-8B-Instruct

vllm unknown on 8x NVIDIA H100 80GB HBM3 (fp16) · seed 0 · run 019e3b30-7ea3-7549-9305-d47c44fa944c · signed

Headline metrics

TTFT P50 (ms) 14.16
TTFT P99 (ms)
Throughput (tok/s)
$/M tokens
J/token
Power avg (W)
Power peak (W)
WER mean
J / audio s

All metrics

chrf_mean 0.7749
chrf_p50 0.7614
chrf_p95 0.9600
n_ok 8.00
n_samples 8.00
ok_rate 1.00
tokens_out_total 151
total_p50_ms 140
ttft_p50_ms 14.16

Provenance

Model revisionunknown00
Providervllm
Hardware fingerprint550474fc9132129654f5d20c316eaffec99a1f67c0cc0b4641e60efb46ea79e7
Datasetbuiltin-flores-mini-en-fr · 25cb8c32504c769c88e3b41ad01bb304ccd100dbc9278b3a41f520053512185e
Timestamp2026-05-18T13:04:54.435962+00:00
Envelope JSONf8c8df08d41d.json

Verify this result

bench verify /bench/leaderboard/envelopes/f8c8df08d41d.json

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