The Myth of Self-Hosted AI: 'Local' Models Still Route Through the Cloud
nomad-nostalgia · reddit · 2026-09-07
- Self-hosting LLMs is trending, but the author argues vendor marketing blurs what "local" really means, since many "self-hosted" setups still require cloud models in the loop.
- Two examples: Superwhisper open-sourced its speech-to-text models (Aug 26) yet local output still needs cloud models for text polishing; Perplexity's Mac hybrid compute (Sep 1) keeps private data local but relies on cloud models for orchestration.
- The author's stance: proprietary cloud models are fine, but "local" should mean end-to-end local inference with no cloud in the loop — not being upsold on something free and required for better outputs.
More from Infra
- DeepSeek V4 Flash at 75% off via Merge Gateway: $0.04/M input tokens through Sept 30 — shensi · 2026-09-07
- Homelab With 4x RTX 4090 Weighs vLLM+P2P Patch vs llama.cpp for Qwen Models — dowitex · 2026-09-07
- Could AI run entirely on your phone? It could upend OpenAI's pricing — kevinsurace · 2026-09-07
- Inference engineering is the underrated AI skill: KV cache, batching and p99 latency explained — techNmak · 2026-09-07
- Polymarket puts 73% odds a US state enacts data center moratorium by end of 2026 — Polymarket · 2026-09-07
- Louisiana taco shop says 40% of monthly business comes from Meta's AI data center — Polymarket · 2026-09-07