Giga Embeddings Instruct vs Giga Embeddings Instruct 10B A1.8B 0826
Giga Embeddings Instruct—
Giga Embeddings Instruct 10B A1.8B 082665.6
| Benchmark | Giga Embeddings Instruct | Giga Embeddings Instruct 10B A1.8B 0826 |
|---|---|---|
| MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages) | — | 65.6 |
Specs
| Spec | Giga Embeddings Instruct | Giga Embeddings Instruct 10B A1.8B 0826 |
|---|---|---|
| Context | 4K | 8K |
| Weights | Open | Open |
| Released | Sep 2025 | Aug 2026 |
| Dimensions | 2,048 | 1,536 |
| Max input | 4K tokens | 8K tokens |
| Languages | en, ru | en, ru |
| Memory (4-bit) | 1.9 GB | 6.3 GB |
| Licence | MIT | MIT |
| Commercial use | Allowed | Allowed |
What breaks when you switch
Giga Embeddings Instruct → Giga Embeddings Instruct 10B A1.8B 0826
Re-embed your whole corpus›
Vectors from different models can't be mixed; dimensions change from 2,048 to 1,536.
Memory (4-bit): 1.9 GB → 6.3 GB
Giga Embeddings Instruct 10B A1.8B 0826 → Giga Embeddings Instruct
Re-embed your whole corpus›
Vectors from different models can't be mixed; dimensions change from 1,536 to 2,048.