Voyage Code 3 measured on 12 of 261 Embeddings benchmarks12
BenchmarkByScoreRank
No benchmark matches
Alternatives?
ModelScoreCostPriceWhat changes
VCheaperVoyage Large 2 Instructunstated+10%-33%$0.12Max input: 32K tokens → 16K tokensModelScoreCostPriceWhat changes
VVoyage Large 2 Instructunstated+10%-33%$0.12Max input: 32K tokens → 16K tokensiE5 Mistral 7B InstructunstatedOpen+6%-88%$0.022ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
KQZhou EmbeddingunstatedOpen+11%——iJasper Token Compression 600MunstatedOpen+8%——KNano Em1 0.6B v2.1unstatedOpen+8%——Max input: 32K tokens → 8K tokensModelScoreCostPriceWhat changes
VVoyage Large 2 Instructunstated+10%-33%$0.12Max input: 32K tokens → 16K tokensVVoyage 3 M Expunstated+4%——VVoyage Law 2unstated+2%-33%$0.12Max input: 32K tokens → 16K tokensVMdbr Leaf MTunstatedOpensame——Max input: 32K tokens → 1K tokensVVoyage Code 2unstatedsame-33%$0.12Max input: 32K tokens → 16K tokensVVoyage Large 2unstatedsame-33%$0.12Max input: 32K tokens → 16K tokensVVoyage 2unstated-1%-44%$0.10Max input: 32K tokens → 4K tokensVVoyage 3 Liteunstated-5%-89%$0.02All Voyage AI (MongoDB) models →History
2 versionsBenchmarks
in Embeddings & RetrievalFAQ
How much does Voyage Code 3 cost?›
$0.18 per 1M tokens.
Is Voyage Code 3 free to use?›
There is no free API; it costs $0.18 per 1M tokens.
How good is Voyage Code 3, and what is it best at?›
#141 of 238 in Embeddings.
Is Voyage Code 3 better than Udever Bloom 1b1?›
On MTEB English v2 · Pair Classification · English, Voyage Code 3 scores 83.1 vs 83.2 — about the same.
When was Voyage Code 3 released? Is there a newer version?›
Released 2024-12-04. It is the current version in its line.
Is Voyage Code 3 open source?›
No, Voyage Code 3 is a closed model available through an API. The closest open-weights option here is QZhou Embedding.
What is Voyage Code 3?›
Voyage Code 3 is a closed embeddings & retrieval model from Voyage AI (MongoDB), released Dec 2024.
What are the best alternatives to Voyage Code 3?›
Voyage Large 2 Instruct (cheaper), EmbeddingGemma 300M (better), QZhou Embedding (open weights).