ColBERT Zero measured on 3 of 261 Embeddings benchmarks3
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
55.4BEIR
Feb 2026released
1K tokensMax input
enLanguages
Alternatives?
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
lmLateOnOpen+4%——Memory (4-bit): 0.1 GB → 0.2 GBlLateOnOpen+3%——Max input: 1K tokens → 0K tokenslmDenseOnOpen+2%——Memory (4-bit): 0.1 GB → 0.2 GBlDenseOnOpensame——lGTE ModernColBERT v1Opensame——Max input: 1K tokens → 0K tokenslColBERT Zero SupervisedOpen-6%——lColBERT Zero UnsupervisedOpen-7%——Max input: 1K tokens → 0K tokenslDenseOn UnsupervisedOpen-11%——All lightonai 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
LicenceApache-2.0free for commercial use
Benchmarks
in Embeddings & RetrievalBEIR · Retrieval · English55.4#41
FAQ
How much does ColBERT Zero cost?›
There is no per-unit API price: self-host only.
Is ColBERT Zero free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is ColBERT Zero, and what is it best at?›
#40 of 194 in Embeddings.
Is ColBERT Zero better than Text Embedding 3 Large?›
On BEIR · Mean over tasks · English, ColBERT Zero scores 55.4 vs 55.4 — about the same.
When was ColBERT Zero released? Is there a newer version?›
Released 2026-02-19. It is the current version in its line.
Is ColBERT Zero open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run ColBERT Zero locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is ColBERT Zero?›
ColBERT Zero is an open-weights embeddings & retrieval model from lightonai, released Feb 2026.
What are the best alternatives to ColBERT Zero?›
Qwen 3 Embedding 8B (better), NV Embed v2 (open weights).