BiUltraVBERT measured on 5 of 203 Multimodal benchmarks5
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
62.5MIEB lite
Oct 2026released
33K tokensMax input
en, es, fr +3Languages
Ranking?
Full ranking →···
42ColSmol 256M66.943BiUltraVBERT62.344E5 V58.1Strong at
Alternatives?
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
DArgus Colqwen3.5 9B v0unstatedOpen+38%——Languages: loses es, fr, de +2+1vVultronRetrieverPrime Qwen3.5 8BOpen+37%——Memory (4-bit): 0.2 GB → 5 GBOOps Colqwen3 4BOpen+36%——Languages: not verified+1ModelScoreCostPriceWhat changes
aColqwen3.5 4.5B v3Open+35%——Languages: loses pt, it+1aColUltraVBERTOpen+24%——All athrael-soju models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · full quality✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceMITfree for commercial use
Benchmarks
in Embeddings & RetrievalFAQ
How much does BiUltraVBERT cost?›
There is no per-unit API price: self-host only.
Is BiUltraVBERT free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is BiUltraVBERT, and what is it best at?›
#53 of 89 in Multimodal.
Is BiUltraVBERT better than ColSmol 256M?›
On ViDoRe v1&v2 · Mean over tasks · multilingual (4 languages), BiUltraVBERT scores 62.3 vs 66.9 — lower.
When was BiUltraVBERT released? Is there a newer version?›
Released 2026-10-18. It is the current version in its line.
Is BiUltraVBERT open source?›
Yes, the weights are open (MIT licence).
Can I run BiUltraVBERT locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is BiUltraVBERT?›
BiUltraVBERT is an open-weights embeddings & retrieval model from athrael-soju, released Oct 2026.
What are the best alternatives to BiUltraVBERT?›
Jina Embeddings v4 (better), Argus Colqwen3.5 9B v0 (open weights).