ColSmol 256M measured on 5 of 203 Multimodal benchmarks5
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
73.4MIEB lite
Jan 2025released
8K tokensMax input
enLanguages
Ranking?
Full ranking →···
41ColSmol 500M71.242ColSmol 256M66.943BiUltraVBERT62.3Strong at
Alternatives?
ModelScoreCostPriceWhat changes
BetterJina Embeddings v4unstatedOpen+21%—$0.05Languages: not verified+1DOpenArgus Colqwen3.5 9B v0unstated+29%——Memory (4-bit): 0.2 GB → 5.3 GBVBetterVoyage Multimodal 3+13%—$0.12Languages: not verifiedvBetterVultronRetrieverPrime Qwen3.5 8BOpen+27%——Memory (4-bit): 0.2 GB → 5 GBOBetterOps Colqwen3 4BOpen+27%——Languages: not verified+1ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Jina Embeddings v4unstatedOpen+21%—$0.05Languages: not verified+1VVoyage Multimodal 3+13%—$0.12Languages: not verifiedDArgus Colqwen3.5 9B v0unstatedOpen+29%——Memory (4-bit): 0.2 GB → 5.3 GBvVultronRetrieverPrime Qwen3.5 8BOpen+27%——Memory (4-bit): 0.2 GB → 5 GBOOps Colqwen3 4BOpen+27%——Languages: not verified+1ModelScoreCostPriceWhat changes
DArgus Colqwen3.5 9B v0unstatedOpen+29%——Memory (4-bit): 0.2 GB → 5.3 GBvVultronRetrieverPrime Qwen3.5 8BOpen+27%——Memory (4-bit): 0.2 GB → 5 GBOOps Colqwen3 4BOpen+27%——Languages: not verified+1
Nemotron Colembed VL 8B v2Open+27%——Languages: not verified+1vVultronRetrieverCore Qwen3.5 4.5BOpen+27%——Memory (4-bit): 0.2 GB → 2.7 GBModelScoreCostPriceWhat changes
vColqwen2.5 v0.2Open+21%——Memory (4-bit): 0.2 GB → 1.8 GBvColqwen2 v1.0Open+19%——Memory (4-bit): 0.2 GB → 1.3 GBvColpali v1.3Open+14%——Memory (4-bit): 0.2 GB → 1.8 GBvColpali v1.2Open+11%——Memory (4-bit): 0.2 GB → 1.8 GBvColpali v1.1Open+9%——Memory (4-bit): 0.2 GB → 1.8 GBvColSmol 500MOpen+6%——Memory (4-bit): 0.2 GB → 0.3 GBAll vidore 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 & RetrievalFAQ
How much does ColSmol 256M cost?›
There is no per-unit API price: self-host only.
Is ColSmol 256M free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is ColSmol 256M, and what is it best at?›
#49 of 89 in Multimodal.
Is ColSmol 256M better than ColSmol 500M?›
On ViDoRe v1&v2 · Mean over tasks · multilingual (4 languages), ColSmol 256M scores 66.9 vs 71.2 — lower.
When was ColSmol 256M released? Is there a newer version?›
Released 2025-01-22. It is the current version in its line.
Is ColSmol 256M open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run ColSmol 256M locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is ColSmol 256M?›
ColSmol 256M is an open-weights embeddings & retrieval model from vidore, released Jan 2025.
What are the best alternatives to ColSmol 256M?›
Jina Embeddings v4 (better), Argus Colqwen3.5 9B v0 (open weights).