Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | intfloatIntfloat: E5-Base-v2Remove × | togethercomputerTogether: Tev1 4B ExperimentalRemove × | googleembeddinggemma-2Remove × |
|---|---|---|---|
| What’s different | The e5-base-v2 embedding model encodes English sentences and paragraphs into a 768-dimensional dense vector space, producing efficient and high-quality semantic embeddings optimized for tasks such as… | Tev1 4B Experimental is an experimental decision model from Together AI, a supervised fine-tune of Qwen3.5-4B trained to choose one option from a structured state, question, and list of 2-24... | EmbeddingGemma 2 is a 740M parameter multimodal embedding model from Google DeepMind that maps text (incl. code), images, video and audio, alone or interleaved, into a single 768-dimensional vector space,… |
| Date | Nov 22, 2023Listed | Sep 30, 2026Listed | Oct 8, 2026Listed |
API prices USD · provider configuration shown below | |||
| Listed rate | $0.005 in / $0.00 out/ 1M tokensOpenRouter · all routes ↗ | $0.042 in / $0.00 out/ 1M tokensOpenRouter · all routes ↗ | See provider quotes |
| Provider endpoints | 1 endpoint → | 1 endpoint → | 0 endpoints → |
Specifications | |||
| Context window | 512 | 32.8K | 8.2K |
| Max output | 460 | 8 | — |
| Parameters | — | 4B | — |
| Open weights | Supported | Supported | Not reported |