Snowflake Arctic Embed M v1.5 vs Snowflake Arctic Embed M v2.0
MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages) Embeddings assign examples to categories, group similar examples, estimate semantic similarity between pairs, classify relationships between pairs, rank relevant documents for queries, find translated sentence pairs, rerank documents according to instructions, assign multiple labels to each example, reorder candidate documents by relevance on MTEB Multilingual v2 · mean over tasks in 1037 languages including 'Auhelawa, Abau, Achi, Achinese, Achuar-Shiwiar, Adzera, Afrikaans, Agarabi; tasks receive equal weight
Snowflake Arctic Embed M v1.542
Snowflake Arctic Embed M v2.0—
| Benchmark | Snowflake Arctic Embed M v1.5 | Snowflake Arctic Embed M v2.0 |
|---|---|---|
| MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages) | 42 | — |
Specs
| Spec | Snowflake Arctic Embed M v1.5 | Snowflake Arctic Embed M v2.0 |
|---|---|---|
| Context | 1K | 8K |
| Weights | Open | Open |
| Released | Jul 2024 | Dec 2024 |
| Dimensions | 768 | 768 |
| Max input | 1K tokens | 8K tokens |
| Languages | — | en, es, Estonian +68 |
| Memory (4-bit) | 0.1 GB | 0.2 GB |
| Licence | Apache-2.0 | Apache-2.0 |
| Commercial use | Allowed | Allowed |
What breaks when you switch
Snowflake Arctic Embed M v1.5 → Snowflake Arctic Embed M v2.0
Re-embed your whole corpus›
Vectors from different models can't be mixed.
Memory (4-bit): 0.1 GB → 0.2 GB
Snowflake Arctic Embed M v2.0 → Snowflake Arctic Embed M v1.5
Re-embed your whole corpus›
Vectors from different models can't be mixed.