mDenseOn measured on 9 of 261 Embeddings benchmarks9
BenchmarkByScoreRank
No benchmark matches
price not verifiedprice
56.7BEIR
Jul 2026released
768Dimensions
8K tokensMax input
en, es, fr +6Languages
Alternatives?
ModelScoreCostPriceWhat changes
Betterpplx Embed v1 0.6BOpen+7%—$0.004Languages: not verified+1
OpenQwen 3 Embedding 8Bunstated+15%—$0.01Languages: not verified+1BetterQwen 3 Embedding 0.6BunstatedOpen+8%—$0.01Languages: not verified+1BetterQwen 3 Embedding 4BunstatedOpen+14%—$0.02Languages: not verified+1Betterpplx Embed v1 4BOpen+11%—$0.03Languages: not verified+1ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
pplx Embed v1 0.6BOpen+7%—$0.004Languages: not verified+1Qwen 3 Embedding 8BunstatedOpen+15%—$0.01Languages: not verified+1Qwen 3 Embedding 0.6BunstatedOpen+8%—$0.01Languages: not verified+1Qwen 3 Embedding 4BunstatedOpen+14%—$0.02Languages: not verified+1pplx Embed v1 4BOpen+11%—$0.03Languages: not verified+1ModelScoreCostPriceWhat changes
Qwen 3 Embedding 8BunstatedOpen+15%—$0.01Languages: not verified+1cC2LLM 7BOpen+15%——Languages: loses es, fr, de +5+1cF2LLM v2 8BunstatedOpen+15%——Memory (4-bit): 0.2 GB → 4.5 GBcF2LLM v2 4BunstatedOpen+15%——Memory (4-bit): 0.2 GB → 2.4 GBQwen 3 Embedding 4BunstatedOpen+14%—$0.02Languages: not verified+1ModelScoreCostPriceWhat changes
lLateOn CodeOpen+6%——Max input: 8K tokens → 2K tokens+1lmLateOnOpen+3%——lLateOn Code EdgeOpen-6%——Max input: 8K tokens → 2K tokens+1lLateOn Code PretrainOpen-13%——Max input: 8K tokens → 2K tokens+1lLateOn Code Edge PretrainOpen-22%——Max input: 8K tokens → 2K tokens+1lDenseOnOpenn/a——Max input: 8K tokens → 1K tokens+1lDenseOn UnsupervisedOpenn/a——Max input: 8K tokens → 1K tokens+1lLateOnOpenn/a——Max input: 8K tokens → 0K tokens+1All 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 · English56.7#29
FAQ
How much does mDenseOn cost?›
There is no per-unit API price: price not verified.
How good is mDenseOn, and what is it best at?›
#29 of 194 in Embeddings.
Is mDenseOn better than BOOM 4B v1?›
On CoIR · Retrieval · multilingual (9 languages), mDenseOn scores 69.3 vs 69.3 — about the same.
When was mDenseOn released? Is there a newer version?›
Released 2026-07-30. It is the current version in its line.
Is mDenseOn open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run mDenseOn locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is mDenseOn?›
mDenseOn is an open-weights embeddings & retrieval model from lightonai, released Jul 2026.
What are the best alternatives to mDenseOn?›
pplx Embed v1 0.6B (better), Qwen 3 Embedding 8B (open weights).