Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | nvidiaNemotron-3-Embed-8BRemove × | thinkingmachinesInkling-SmallRemove × | googleembeddinggemma-2Remove × |
|---|---|---|---|
| What’s different | Nemotron-3-Embed-8B is a multilingual text embedding model from NVIDIA, based on Ministral-3-8B, that maps text into 4096-dimensional dense vectors for retrieval and semantic similarity. Spanning 34… | Inkling-Small is a Mixture-of-Experts transformer with 276B total parameters, 12B active, trained on NVIDIA GB300 NVL72 systems. Like Inkling, it features native reasoning over audio and images, variable… | 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 | Jul 16, 2026Listed | Jul 27, 2026Listed | Oct 8, 2026Listed |
API prices USD · provider configuration shown below | |||
| Listed rate | See provider quotes | $0.45 in / $1.20 out/ 1M tokensdeepinfra · via Hugging Face · all routes ↗ | See provider quotes |
| Provider endpoints | 0 endpoints → | 4 endpoints → | 0 endpoints → |
Specifications | |||
| Context window | 32.8K | 524.3K | 8.2K |
| Max output | — | 262.1K | — |
| Parameters | 8B | — | — |
| Tool calling | Not reported | Supported | Not reported |
| Structured output | Not reported | Supported | Not reported |
| Image input | Not reported | Supported | Not reported |
| Audio input | Not reported | Supported | Not reported |
| Reasoning | Not reported | Supported | Not reported |
| Prompt caching | Not reported | Supported | Not reported |
| Open weights | Not reported | Supported | Not reported |