MoEdernColBERT-115M matches Qwen3-Embedding-0.6B at 8x throughput
tomaarsen · x · 2026-10-26
tomaarsen spotlights VAGOsolutions' MoEdernColBERT line (1.5M/27M/115M), the smallest MoE embedding retrievers he's seen. Per the team's chart, the 115M model beats Qwen3-Embedding-0.6B on 9 of 11 benchmarks (English NanoBEIR the exception, 0.643 vs 0.648), leads German tasks like GermanQuAD (0.989) and LegalQuAD (0.851) by wide margins, outperforms the 2.7x-larger granite-311m-multilingual-r2 across all 11, and encodes at 1,132 docs/s — 8x Qwen3-0.6B and at least 2.5x every compared model.
Related event: VAGOsolutions releases tiny 1.5M-115M MoE ColBERT retrieval models(10 posts)→
More from Models
- Gemma 4 26B-A4B runs offline on a 16 GB Mac: 33/40 on TypeScript tasks at 31 tok/s — Mission-Sir-1614 · 2026-10-26
- Claude safety classifier blocks SS7 attack analysis, dev pivots to abliterated GLM — l4rz · 2026-10-26
- Sakana AI to pretrain first domestic model Sakana Sawara for defense and cybersecurity by 2027 — tkasasagi · 2026-10-26
- 31M-param English TTS model phonon-31m-en trends on Hugging Face — gradium · 2026-10-26
- NVIDIA tutorial: post-training Nemotron for Europe with Persona datasets and LoRA — NVIDIA Developer · 2026-10-26
- Opus 5 hailed as the 'most emo, greatest artist' model by AI community users — repligate · 2026-10-26