blob: 53d46dff57535869082a87337a48540840ae03f2 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
|
# Generate Embeddings
`POST /api/embed` — Creates vector embeddings representing the input text.
**Server:** `http://localhost:11434`
Embeddings turn text into numeric vectors for semantic search, retrieval, and RAG pipelines. Vector length depends on the model (typically 384–1024 dimensions). Vectors are **L2-normalized** (unit-length).
## Recommended Models
- `embeddinggemma`
- `qwen3-embedding`
- `all-minilm`
## Request
| Field | Type | Required | Description |
|---|---|---|---|
| `model` | string | yes | Model name (e.g. `"embeddinggemma"`) |
| `input` | string \| string[] | yes | Text or array of texts to embed |
| `truncate` | boolean | no | Truncate inputs exceeding context window (default: `true`). If `false`, returns an error. |
| `dimensions` | integer | no | Number of dimensions for the embedding vectors |
| `keep_alive` | string | no | Model keep-alive duration |
| `options` | ModelOptions | no | Runtime options (see generate.md) |
## Response
| Field | Type | Description |
|---|---|---|
| `model` | string | Model that produced the embeddings |
| `embeddings` | number[][] | Array of embedding vectors (one per input) |
| `total_duration` | integer | Total time (nanoseconds) |
| `load_duration` | integer | Model load time (nanoseconds) |
| `prompt_eval_count` | integer | Number of input tokens processed |
## Examples
### Single input
```bash
curl http://localhost:11434/api/embed -d '{
"model": "embeddinggemma",
"input": "Why is the sky blue?"
}'
```
### Multiple inputs (batch)
```bash
curl http://localhost:11434/api/embed -d '{
"model": "embeddinggemma",
"input": [
"Why is the sky blue?",
"Why is the grass green?"
]
}'
```
### Custom dimensions
```bash
curl http://localhost:11434/api/embed -d '{
"model": "embeddinggemma",
"input": "Generate embeddings for this text",
"dimensions": 128
}'
```
## Tips
- Use **cosine similarity** for most semantic search use cases.
- Use the **same embedding model** for both indexing and querying.
- Batch multiple strings in one request for efficiency.
|
