# Nemotron 3.5 Embed 8B

> NVIDIA Nemotron 3.5 Embed 8B is a 8.38B-parameter multimodal and multilingual retrieval embedding model. It embeds text (34 evaluated languages) and, on the model card, document-page images, into one 4096-dimension vector for semantic search and RAG; vectors can be truncated and re-normalized. 32,768-token context. On AnyRouter's embeddings endpoint the input is text. It is the teacher model for the 1B variant.

**ID**: `nvidia/nemotron-3.5-embed-8b`  
**Creator**: nvidia  
**Category**: embedding  
**Context**: 33K tokens  
**Input**: $0  
**Output**: $0  
**Released**: 2026-10-20  
**Web page**: https://anyrouter.dev/model/nvidia/nemotron-3.5-embed-8b

**Input modalities**: text  
**Output modalities**: embeddings  

**Tokenizer**: Mistral

**Capabilities**: embedding

**Supported parameters**: input, model, encoding_format

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/embeddings`

```bash
curl https://anyrouter.dev/api/v1/embeddings \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "nvidia/nemotron-3.5-embed-8b",
  "input": "The quick brown fox jumps over the lazy dog"
}'
```

Pass `input` as a string or an array of strings to embed a batch in one call.

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| nvidia | $0 | $0 |
| nvidia-byok | $0 | $0 |

## Source

- Upstream docs: https://build.nvidia.com/nvidia/nemotron-3.5-embed-8b