# MTEB SciFact 28 modelli con un valore. In testa: Qwen3 Embedding 8B con 78,5. | Scheda del benchmark | | |---|---| | Compito | Retrieval (en) | | Metrica | nDCG@10 | | Direzione | ↑ più alto è meglio | | Limitazione | Ein Wert gilt fuer die gemessene Modellrevision und die volle Vektorlaenge; gekuerzte Laengen sind nicht gemessen (R-MESS). | | Fonte | [huggingface.co](https://huggingface.co/spaces/mteb/leaderboard) | | Licenza | CC0 — MTEB-Ergebnisse (embeddings-benchmark/results) | ## Classifica (28) | Modello | Posizione | Valore | Sviluppatore | Trattamento nell'UE: Input | Trattamento nell'UE: Output | tutti i fornitori: Input | tutti i fornitori: Output | Pesi aperti | Memoria GPU (4 bit, stimata) | |---|---:|---:|---|---:|---:|---:|---:|---|---:| | [Qwen3 Embedding 8B](https://models.searchit-agent.ai/it/modelli/qwen3-embedding-8b/) | 1 | **78,5** | Alibaba Cloud | 0,009 € | – | 0,009 € | – | ✓ sì | [da 9 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [E5-Mistral-7B](https://models.searchit-agent.ai/it/modelli/e5-mistral-7b/) | 2 | **76,4** | Microsoft | 0,02 € | – | 0,02 € | – | ✓ sì | [da 8 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [bge-large-en-v1.5](https://models.searchit-agent.ai/it/modelli/bge-large-en-v1-5/) | 3 | **74,6** | BAAI | – | – | 0,009 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [bge-base-en-v1.5](https://models.searchit-agent.ai/it/modelli/bge-base-en-v1-5/) | 4 | **74,3** | BAAI | – | – | 0,0045 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [mxbai-embed-large-v1](https://models.searchit-agent.ai/it/modelli/mxbai-embed-large-v1/) | 5 | **73,9** | Mixedbread | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [snowflake-arctic-embed-l](https://models.searchit-agent.ai/it/modelli/snowflake-arctic-embed-l/) | 6 | **73,8** | Snowflake | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [gte-multilingual-base](https://models.searchit-agent.ai/it/modelli/gte-multilingual-base/) | 7 | **73,4** | Alibaba Cloud | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [bekko-embedding-v1-a25m](https://models.searchit-agent.ai/it/modelli/bekko-embedding-v1-a25m/) | 8 | **72,3** | hotchpotch | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [snowflake-arctic-embed-m-v2.0](https://models.searchit-agent.ai/it/modelli/snowflake-arctic-embed-m-v2-0/) | 8 | **72,3** | Snowflake | – | – | 0,036 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [BGE Multilingual Gemma2](https://models.searchit-agent.ai/it/modelli/bge-multilingual-gemma2/) | 10 | **72** | BAAI | 0,01 € | – | 0,01 € | – | ✓ sì | [da 10 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [SNOWFLAKE ARCTIC-EMBED-M-V1.5](https://models.searchit-agent.ai/it/modelli/snowflake-arctic-embed-m-v1-5/) | 11 | **71,84** | Snowflake | – | – | 0,022 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [multilingual-e5-large-instruct](https://models.searchit-agent.ai/it/modelli/multilingual-e5-large-instruct/) | 12 | **71,6** | Microsoft | 0,03 € | 0,02 € | 0,009 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [snowflake-arctic-embed-l-v2.0](https://models.searchit-agent.ai/it/modelli/snowflake-arctic-embed-l-v2-0/) | 13 | **70,9** | Snowflake | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [granite-embedding-311m-multilingual-r2](https://models.searchit-agent.ai/it/modelli/granite-embedding-311m-multilingual-r2/) | 14 | **70,8** | IBM | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [Amazon Titan Text Embeddings V2](https://models.searchit-agent.ai/it/modelli/titan-embed-text-v2/) | 15 | **70,5** | Amazon | – | – | – | – | – no | – | | [nomic-embed-text-v1.5](https://models.searchit-agent.ai/it/modelli/nomic-embed-text-v1-5/) | 16 | **70,3** | Nomic AI | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [intfloat/multilingual-e5-large](https://models.searchit-agent.ai/it/modelli/multilingual-e5-large/) | 17 | **70,2** | Microsoft | 0,0082 € | – | 0,0082 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [bekko-embedding-v1-a8m](https://models.searchit-agent.ai/it/modelli/bekko-embedding-v1-a8m/) | 18 | **70** | hotchpotch | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [cohere-embed-multilingual-v3.0](https://models.searchit-agent.ai/it/modelli/cohere-embed-multilingual-v3-0/) | 18 | **70** | Cohere | – | – | 0,09 € | – | – no | – | | [snowflake-arctic-embed-m-long](https://models.searchit-agent.ai/it/modelli/snowflake-arctic-embed-m-long/) | 18 | **70** | Snowflake | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [Qwen3-Embedding-0.6B](https://models.searchit-agent.ai/it/modelli/qwen3-embedding-0-6b/) | 21 | **69,7** | Alibaba Cloud | – | – | 0,011 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [granite-embedding-97m-multilingual-r2](https://models.searchit-agent.ai/it/modelli/granite-embedding-97m-multilingual-r2/) | 22 | **68,7** | IBM | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [multilingual-e5-small](https://models.searchit-agent.ai/it/modelli/multilingual-e5-small/) | 23 | **67,58** | Google | – | – | 0,0067 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [granite-embedding-278m-multilingual](https://models.searchit-agent.ai/it/modelli/granite-embedding-278m-multilingual/) | 24 | **65,8** | IBM | – | – | – | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [all-mpnet-base-v2](https://models.searchit-agent.ai/it/modelli/all-mpnet-base-v2/) | 25 | **65,57** | – | – | – | 0,0045 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [all-MiniLM-L6-v2](https://models.searchit-agent.ai/it/modelli/all-minilm-l6-v2/) | 26 | **64,5** | – | – | – | 0,0045 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [BGE-M3](https://models.searchit-agent.ai/it/modelli/bge-m3/) | 27 | **64,4** | BAAI | 0,0082 € | – | 0,0082 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | | [All MiniLM L12 v2](https://models.searchit-agent.ai/it/modelli/all-minilm-l12-v2/) | 28 | **62,64** | – | – | – | 0,0045 € | – | ✓ sì | [da 5 GB](https://models.searchit-agent.ai/it/self-hosting/#tier-24-gb) | _Valori uguali = stessa posizione. Con i filtri la posizione resta quella nella lista completa. La memoria GPU «da …» è un limite inferiore (parametri × 0,55 GB + 4 GB). Prezzi: € per 1 mln di token. Fonte: pagine prezzi dei fornitori, lette e verificate da searchit. Al 7 nov 2026._ _Valori: MTEB SciFact, CC0 — MTEB-Ergebnisse (embeddings-benchmark/results). Ripresi invariati dalla fonte._ --- Prezzi al 7 nov 2026. Prezzi di listino dei fornitori, senza garanzia.