LateOn Code measured on 6 of 261 Embeddings benchmarks6
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
73.4CoIR
Feb 2026released
2K tokensMax input
enLanguages
Alternatives?
ModelScoreCostPriceWhat changes
BetterQwen 3 Embedding 8BunstatedOpen+9%—$0.01Languages: not verified+1BetterQwen 3 Embedding 4BunstatedOpen+8%—$0.02Languages: not verified+1
Betterpplx Embed v1 4BOpen+5%—$0.03Languages: not verified+1VBetterVoyage Code 3unstated+7%—$0.18Languages: not verifiedcBetterF2LLM v2 14BunstatedOpen+9%——Memory (4-bit): 0.1 GB → 8.4 GBModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Qwen 3 Embedding 8BunstatedOpen+9%—$0.01Languages: not verified+1Qwen 3 Embedding 4BunstatedOpen+8%—$0.02Languages: not verified+1pplx Embed v1 4BOpen+5%—$0.03Languages: not verified+1VVoyage Code 3unstated+7%—$0.18Languages: not verifiedcF2LLM v2 14BunstatedOpen+9%——Memory (4-bit): 0.1 GB → 8.4 GBModelScoreCostPriceWhat changes
Qwen 3 Embedding 8BunstatedOpen+9%—$0.01Languages: not verified+1cC2LLM 7BOpen+8%——Memory (4-bit): 0.1 GB → 4.6 GBcF2LLM v2 8BunstatedOpen+8%——Memory (4-bit): 0.1 GB → 4.5 GBcF2LLM v2 4BunstatedOpen+8%——Memory (4-bit): 0.1 GB → 2.4 GBQwen 3 Embedding 4BunstatedOpen+8%—$0.02Languages: not verified+1ModelScoreCostPriceWhat changes
lmLateOnOpen-3%——Memory (4-bit): 0.1 GB → 0.2 GBlmDenseOnOpen-6%——Memory (4-bit): 0.1 GB → 0.2 GBlLateOn Code EdgeOpen-12%——lLateOn Code PretrainOpen-18%——lLateOn Code Edge PretrainOpen-26%——lDenseOnOpenn/a——Max input: 2K tokens → 1K tokenslDenseOn UnsupervisedOpenn/a——Max input: 2K tokens → 1K tokenslLateOnOpenn/a——Max input: 2K tokens → 0K tokensAll 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 & RetrievalFAQ
How much does LateOn Code cost?›
There is no per-unit API price: self-host only.
Is LateOn Code free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is LateOn Code, and what is it best at?›
#22 of 58 in Embeddings.
Is LateOn Code better than Gemini Embedding 001?›
On CoIR · Retrieval · multilingual (9 languages), LateOn Code scores 73.4 vs 73.9 — about the same.
When was LateOn Code released? Is there a newer version?›
Released 2026-02-12. It is the current version in its line.
Is LateOn Code open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run LateOn Code locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is LateOn Code?›
LateOn Code is an open-weights embeddings & retrieval model from lightonai, released Feb 2026.
What are the best alternatives to LateOn Code?›
Qwen 3 Embedding 8B (better).