GroqCloud supported-models page (model ids, vendor speed figures, per-1M prices)

Fetched 2026-10-14. Production rows: openai/gpt-oss-120b $0.15/$0.60 at vendor-stated ~500 t/s, 131072 ctx, max completion 65536; openai/gpt-oss-20b $0.075/$0.30 at ~1000 t/s. llama-3.1-8b-instant (560 t/s) and llama-3.3-70b-versatile (280 t/s, 131072 ctx) are both Enterprise-gated: price column reads literally 'ContactSales'. Preview: qwen/qwen3.8-27b $0.80/$4.00 at 450 t/s; openai/gpt-oss-safeguard-20b $0.075/$0.30. Whisper priced per hour, Canopy Orpheus per 1M characters. Supersedes the 2026-10-04 fetch of the same page (rent-src-groq-model-pricing): gpt-oss prices unchanged, Llama still gated.