Airbnb fine-tuned Gemma 4 E2B to match frontier quality at 1% of the cost
DynamicWebPaige · x · 2026-11-07
Airbnb Engineering fine-tuned Google's Gemma 4 E2B on a single GPU with vLLM to serve production classifiers. Results: 35ms median latency (85ms p99), 0.902 F1 matching frontier models, 1% of the cost at scale. Using their internal GLOWC framework and logprob-based thresholding, the fine-tuned E2B beats GPT/Claude on the quality-latency frontier — a strong case for fine-tuned small models over long-prompt frontier APIs in high-volume workloads.
More from Infra
- Tesla targets chip fab inefficiency: wafers spend as little as 20% of time in useful processing — XFreeze · 2026-11-07
- Redditor runs Qwen Flash Next locally at 24 tok/s on 64GB MacBook via Strata SSD offloading — New-Pressure-6932 · 2026-11-07
- Reactor co-founder publishes 'The real-time manifesto' on serving real-time AI at scale — smtabatabaie · 2026-11-07
- Clairvoyance's new hivemind feature lets you share your GPU with anyone — draginol · 2026-11-07
- zml/llmd inference server runs on NVIDIA, AMD, Intel, TPU and Metal with images as small as 142MB — siwu · 2026-11-07
- Sequoia's Maguire models SpaceX taking inference compute share as neoclouds struggle — shaunmmaguire · 2026-11-07