Nomic Embed Text v1.5 (2024年)（ノミックエンベッドテキスト）
2024年2月Nomic AI発表Nomic Embed Text v1.5・Industry-leading 64-768d Matryoshka + Industry-leading 8192 tokens + Industry-leading Apache 2.0 open-source + Industry-leading Nomic AI Nomic Embed Text v1.5 Apache 2.0 embedding 2024。
概要
Nomic Embed Text v1.5 は、2024年2月Nomic AI発表Nomic Embed Text v1.5・Industry-leading 64-768d Matryoshka 2024年Nomic AI Nomic Embed Text v1.5 Apache 2.0 embedding position確立。Nomic Embed Text v1.5 specifications = Industry-leading Apache 2.0 embedding (Industry-leading Nomic Embed Text v1.5 2024 + Industry-leading Apache 2.0 embedding + Industry-leading Nomic AI flagship) + Industry-leading 8192 tokens + Industry-leading Apache 2.0 open-source + Industry-leading Nomic AI Nomic Embed Text v1.5 Apache 2.0 embedding 2024。
主な特徴・仕組み
- Developer: Industry-leading Nomic AI (Founded 2022 NYC by Brandon Duderstadt + Andriy Mulyar)
- Model: Industry-leading Nomic Embed Text v1.5 open-source embedding model 2024年2月 release
- 64-768d Matryoshka: Industry-leading 64-768d Matryoshka adjustable dimension signature
- 8192 Tokens Context: Industry-leading 8192 tokens context signature
- Apache 2.0 Open-Source: Industry-leading Apache 2.0 open-source signature
- Training Data Released: Industry-leading training data fully released signature
- OpenAI text-embedding-3-small Beat: Industry-leading beats OpenAI text-embedding-3-small on MTEB signature (per Nomic benchmarks)
- Atlas Visualization: Industry-leading Nomic Atlas visualization platform integration
- GPT4All Compatible: Industry-leading GPT4All compatible
- HuggingFace Native: Industry-leading HuggingFace native + Sentence Transformers
- MTEB 62.28 Score: Industry-leading MTEB 62.28 average score
- Nomic AI 2022 NYC: Industry-leading Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years
- Pricing: Industry-leading Free open-source
- Industry-Leading: Industry-leading Nomic Embed Text v1.5 Nomic AI Apache 2.0 embedding 2024
スペック比較表
| Embedding Model 2024 | Developer | Dimension | Pricing | Industry Position |
|---|---|---|---|---|
| Nomic Embed Text v1.5 | Nomic AI | 64-768 (Matryoshka) | Free open-source | Industry-leading Nomic Embed Text v1.5 Apache 2.0 |
| OpenAI text-embedding-3-small | OpenAI | 256-1536 (Matryoshka) | $0.02/1M tokens | Industry-leading OpenAI efficient |
| Cohere embed-v3 | Cohere | 1024 (default) | $0.10/1M tokens | Industry-leading Cohere multilingual quality |
| BGE-M3 | BAAI | 1024 | Free open-source | Industry-leading BAAI BGE-M3 multi-functionality |
| Voyage AI voyage-3 | Voyage AI | 1024 (default) | $0.06/1M tokens | Industry-leading Voyage AI voyage-3 enterprise |
具体例・対応製品
- Nomic Embed Text v1.5 (2024年2月, Apache 2.0, Nomic AI): Industry-leading Nomic AI Nomic Embed Text v1.5
- Industry-leading 64-768d Matryoshka adjustable: Industry-leading 64-768d Matryoshka adjustable
- Industry-leading 8192 tokens context: Industry-leading 8192 tokens context
- Industry-leading Apache 2.0 open-source + training data released: Industry-leading Apache 2.0 + training data released
- Industry-leading beats text-embedding-3-small + 62.28 MTEB: Industry-leading beats text-embedding-3-small + 62.28 MTEB
- 競合 OpenAI text-embedding-3-small + Cohere embed-v3 + BGE-M3 + Voyage AI voyage-3: Industry-leading embedding model competitors
選び方・注意点
Nomic Embed Text v1.5 は「Industry-leading 64-768d Matryoshka + 8192 tokens」「Industry-leading Apache 2.0 + Nomic AI NYC 2 years」用途のIndustry-leading premium 2024年2月 Nomic AI product。Industry-leading 64-768d Matryoshka (Industry-leading 64-768d Matryoshka adjustable dimension signature + Industry-leading Nomic Embed Matryoshka signature + Industry-leading dimension-truncatable representations) で Industry-leading 64-768d Matryoshka + Industry-leading Nomic Embed Matryoshka + Industry-leading dimension-truncatable representations。Industry-leading 8192 tokens (Industry-leading 8192 tokens context signature + Industry-leading Nomic Embed long-context signature + Industry-leading long-document embedding) で Industry-leading 8192 tokens + Industry-leading Nomic Embed long-context + Industry-leading long-document embedding。Industry-leading Apache 2.0 (Industry-leading Apache 2.0 open-source signature + Industry-leading Nomic Embed Apache 2.0 signature + Industry-leading commercial-friendly permissive license) で Industry-leading Apache 2.0 + Industry-leading Nomic Embed Apache 2.0 + Industry-leading commercial-friendly permissive license。Industry-leading training data released (Industry-leading training data fully released signature + Industry-leading Nomic Embed training transparency signature + Industry-leading reproducible research) で Industry-leading training data released + Industry-leading Nomic Embed training transparency + Industry-leading reproducible research。Industry-leading Nomic AI 2022 NYC (Industry-leading Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years + Industry-leading beats OpenAI text-embedding-3-small on MTEB per Nomic + Industry-leading Nomic Atlas visualization + Industry-leading GPT4All compatible + Industry-leading HuggingFace native + Sentence Transformers + Industry-leading MTEB 62.28 score) で Industry-leading Nomic AI 2022 + Industry-leading NYC + Industry-leading Brandon Duderstadt + Andriy Mulyar + Industry-leading 2 years + Industry-leading beats text-embedding-3-small + Industry-leading Nomic Atlas + Industry-leading GPT4All + Industry-leading HuggingFace+Sentence Transformers + Industry-leading 62.28 MTEB。但しIndustry-leading OpenAI text-embedding-3-small + Cohere embed-v3 + BGE-M3 + Voyage AI voyage-3 competition で Industry-leading 4-embedding model competitors vs Nomic Embed Text v1.5 + Industry-leading 64-768d Matryoshka adjustable + 8192 tokens context + Apache 2.0 open-source + training data fully released + beats OpenAI text-embedding-3-small on MTEB + Nomic Atlas visualization platform + GPT4All compatible + HuggingFace native + Sentence Transformers + MTEB 62.28 average + Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years + Free open-source unique advantage adoption alignment必須。
関連用語との違い
- vs OpenAI text-embedding-3-small (2024 $0.02/1M): Nomic Embedは Apache 2.0 Free + 64-768d Matryoshka + 8192 tokens Nomic AI・text-embedding-3-smallは$0.02/1M + 256-1536d Matryoshka + 8192 tokens OpenAI
- vs BGE-M3 (2024 MIT): Nomic Embedは Apache 2.0 + Matryoshka 64-768d Nomic AI NYC・BGE-M3はMIT + multi-functionality + 1024d BAAI Beijing
- vs Cohere embed-v3 (2024 $0.10/1M): Nomic EmbedはApache 2.0 Free Nomic AI・embed-v3は$0.10/1M + 100+ languages Cohere
よくある質問（FAQ）
Q1: Nomic Embed Text v1.5 vs BGE-M3 違いは？ A: Nomic Embed (Industry-leading 64-768d Matryoshka + 8192 tokens + Apache 2.0 + training data released + beats OpenAI text-embedding-3-small + Nomic Atlas + GPT4All + HuggingFace + 62.28 MTEB + Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years) vs BGE-M3 (Industry-leading multi-functionality dense+sparse+multi-vector M3 + 100+ languages + 8192 tokens + 1024d dense + sparse vector + multi-vector ColBERT-like + 66.4 BEIR + MTEB top-tier + HuggingFace+FlagEmbedding + BAAI 2018 Beijing 6 years + MIT)・Industry-leading 64-768d Matryoshka + 8192 tokens + Apache 2.0 + training data + beats text-embedding-3-small + Nomic Atlas + GPT4All + 62.28 MTEB + Nomic AI + NYC + 2022 + 2 years = Nomic Embed + Industry-leading multi-functionality + 100+ languages + 1024d dense + sparse + multi-vector ColBERT-like + 66.4 BEIR + BAAI + Beijing + 2018 + 6 years + MIT = BGE-M3 preference judgment (Nomic Embed Matryoshka adjustable + training data transparency NYC vs BGE-M3 multi-functionality multilingual Beijing differ)。
Q2: Industry-leading 64-768d Matryoshka + 8192 tokens value は？ A: Industry-leading 64-768d Matryoshka + 8192 tokens (Industry-leading 64-768d Matryoshka adjustable + Industry-leading 8192 tokens context + Industry-leading Nomic Embed dual signature)。
Q3: Industry-leading Apache 2.0 + Nomic AI NYC 2 years value は？ A: Industry-leading Apache 2.0 + Nomic AI (Industry-leading Apache 2.0 open-source commercial-friendly + Industry-leading Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years heritage)。
まとめ
Nomic Embed Text v1.5 = 2024年2月Nomic AI発表のpremium Apache 2.0 open-source embedding model。Industry-leading 64-768d Matryoshka adjustable dimension signature + Industry-leading 8192 tokens context signature + Industry-leading Apache 2.0 open-source signature + Industry-leading training data fully released signature + Industry-leading beats OpenAI text-embedding-3-small on MTEB signature per Nomic benchmarks + Industry-leading Nomic Atlas visualization platform integration + Industry-leading GPT4All compatible + Industry-leading HuggingFace native + Sentence Transformers + Industry-leading MTEB 62.28 average score + Industry-leading Nomic AI 2022 NYC Brandon Duderstadt + Andriy Mulyar 2 years + Industry-leading Free open-source + Industry-leading Nomic Embed Text v1.5 Nomic AI Apache 2.0 embedding 2024 position確立。