Skip to content

Commit 266d6ce

Browse files
authored
Vectorize: clarify vector dimension storage precision (cloudflare#23445)
1 parent 9f752a8 commit 266d6ce

File tree

2 files changed

+5
-2
lines changed

2 files changed

+5
-2
lines changed

src/content/docs/vectorize/best-practices/insert-vectors.mdx

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -21,14 +21,16 @@ Use the upsert operation if you want to overwrite the vector value for a vector
2121

2222
## Supported vector formats
2323

24-
Vectorize supports vectors in three formats:
24+
Vectorize supports the insert/upsert of vectors in three formats:
2525

2626
- An array of floating point numbers (converted into a JavaScript `number[]` array).
2727
- A [Float32Array](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Float32Array)
2828
- A [Float64Array](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Float64Array)
2929

3030
In most cases, a `number[]` array is the easiest when dealing with other APIs, and is the return type of most machine-learning APIs.
3131

32+
Vectorize stores and restitutes vector dimensions as Float32; vector dimensions provided as Float64 will be converted to Float32 before being stored.
33+
3234
## Metadata
3335

3436
Metadata is an optional set of key-value pairs that can be attached to a vector on insert or upsert, and allows you to embed or co-locate data about the vector itself.

src/content/docs/vectorize/platform/limits.mdx

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -10,7 +10,8 @@ The following limits apply to accounts, indexes and vectors (as specified):
1010
| Feature | Current Limit |
1111
| ------------------------------------------------------------- | ----------------------------------- |
1212
| Indexes per account | 50,000 (Workers Paid) / 100 (Free) |
13-
| Maximum dimensions per vector | 1536 dimensions |
13+
| Maximum dimensions per vector | 1536 dimensions, 32 bits precision |
14+
| Precision per vector dimension | 32 bits (float32) |
1415
| Maximum vector ID length | 64 bytes |
1516
| Metadata per vector | 10KiB |
1617
| Maximum returned results (`topK`) with values or metadata | 20 |

0 commit comments

Comments
 (0)