OpenRouter provider table for meta-llama/llama-3.3-70b-instruct, fifth re-read 2026-10-17 (per-provider price, P50 latency, throughput, uptime)

Fetched later on 2026-10-17 than svc-src-openrouter-provider-table-2026-10-17-r4, same URL. Per-provider rows for Llama 3.3 70B Instruct, P50 over OpenRouter's own routed traffic. DeepInfra (Turbo) $0.100/$0.320, 0.65s latency, 16 tps, 99.85% uptime (was 1.71s / 6 tps / 99.92% on the r4 read). NovitaAI $0.135/$0.400, 1.30s, 35 tps, 99.35% (was 1.05s / 41 tps / 99.29%). AkashML $0.200/$0.520 (cache read $0.100), 0.68s, 31 tps, 99.87%. Parasail $0.220/$0.500 (cache read $0.110), 0.87s, 46 tps, 100.00%. Cloudflare $0.293/$2.253, 0.82s, 41 tps, 99.47%. SambaNova Turbo list $0.60/$1.20 struck through next to a 25% off badge rendering $0.450/$0.900, 0.89s, 86 tps, 99.87% (was 2.58s / 162 tps / 99.86%). Groq $0.590/$0.790 (cache read $0.295), 0.52s, 191 tps, 100.00% (was 0.51s / 198 tps). CoreWeave $0.710/$0.710 (cache read $0.710), 0.13s, 85 tps, 100.00% (was 0.11s / 88 tps). Google Vertex (US flag) $0.720/$0.720, 2.26s, 103 tps; a second Vertex row at the same price has no latency cell (r4's US row was 1.22s / 96 tps and the second row was 1.40s / 96 tps). Together $1.040/$1.040, 0.94s, 28 tps, 96.72% (was 0.55s / 41 tps / 96.68%). Page headline price is still the floor route $0.10/$0.32. Best-across-providers throughput widget reads 191 tok/s (Groq); best latency widget reads 0.13s (CoreWeave). List prices on every row are unchanged from the r4 read, including the SambaNova struck-through discount. These are OpenRouter-measured P50 numbers over its own traffic, not vendor claims and not the Artificial Analysis long-prompt medians (re-fetched the same pass: Groq still 305.7 t/s, SambaNova 286.3, Vertex 153.6, DeepInfra 14.9).