Yuma backs Bittensor subnet Fugal claiming GPT-5.5-level results at 1/6 the cost
markjeffrey · x · 2026-11-05
Yuma Group has admitted fugalai (Bittensor subnet SN70) into its Accelerator. Fugal builds an intelligent dispatcher for inference: a lightweight router analyzes each query, picks the cheapest model likely to answer correctly, and calls a flagship model only when the task demands it.
Fugal claims its own measurements achieve GPT-5.5-level results at one-sixth of the cost (unverified by third parties). On Bittensor, SN70 miners compete to make better routing decisions rather than train another LLM, continuously improving the quality/cost tradeoff.
More from Venture
- Acquisition day: Pierre to OpenAI, Conductor to Meta, HumanLayer to Factory — evilrabbit_ · 2026-11-05
- Hugging Face on who profits from the open source AI boom amid Chinese model competition — TechCrunch AI · 2026-11-05
- Small models ace payout fraud review in Kill Bill — same answers, 106x cost gap vs Claude — m4rkGS · 2026-11-05
- Lenny's Newsletter goes print: 8 AI-era career posts as a $29 magazine — lennysan · 2026-11-05
- Hustle Fund reviewed 2,718 pitches in October and funded just 8 — dunkhippo33 · 2026-11-05
- Accounting can't be one-shotted: Accrual on rebuilding tax workflows with AI — marcbhargava · 2026-11-05