New Bittensor inference subnet Terza undercuts official model prices by up to 96%
markjeffrey · x · 2026-10-31
Terza Inference (Bittensor Subnet 103) is in early access: one API key routing requests across independent miners, billed on verified delivery. Claimed per-1M-token pricing vs. official: GPT-6.1 Sol and GPT-6 Astra 96% cheaper, Claude Sonnet 5.5 and Opus 5.5 89% cheaper, Gemini 3.8 Flash 91% cheaper, Kimi K3 64% cheaper, GLM-5.3 40% cheaper — also undercutting rival subnets Chutes, SayGM, and engy. The poster hasn't tested it.
More from Infra
- AI scarcity moved to atoms: memory, GPU and wafer margins hit 58%-80% — sujingshen · 2026-10-31
- Anthropic's S-1 risk factors: compute tops the list at 13.4%, edging out regulation — ayushtweetshere · 2026-10-31
- TU Eindhoven shows glass interposer with BGA below and MPO optical port on top — jwt0625 · 2026-10-31
- AI Boom Triggers Massive PCB Shortage With Orders Booked Until 2030 — chemist_slime · 2026-10-31
- Beff Jezos doubles down on Cerebras: GPU backends feel 'stupid' for frontier-scale fast inference — beffjezos · 2026-10-31
- Intel Arc B580/B70 ComfyUI Benchmarks: omnixpu Beats Native pytorchxpu — Willing-Grade5566 · 2026-10-31