?crossposting #workers-ai

Please do not post your question in multiple channels/post it multiple times per the rules at #

welcome-and-rules. It creates confusion for people trying to help you and doesn't get your issue or question solved any faster.

Isaac McFadyenOP•4/20/25, 8:34 PM

For future use there's a dedicated #autorag channel

Shadow•4/21/25, 11:51 AM

Hey,
I’m encountering a persistent issue when querying a Cloudflare vector index using env.VECTORS.query() in a worker. Despite passing a valid 768-dimensional vector, I consistently receive the error.

VECTOR_QUERY_ERROR (code = 40006): invalid query vector, expected 768 dimensions, and got 0 dimensions

VECTOR_QUERY_ERROR (code = 40006): invalid query vector, expected 768 dimensions, and got 0 dimensions

I have validated the entries with:

console.log(“Vector length:”, queryVectorRaw.length); // 768

console.log(“Vector length:”, queryVectorRaw.length); // 768

console.log(“typeof vector[0]:”, typeof queryVectorRaw[0]); // “number”

console.log(“typeof vector[0]:”, typeof queryVectorRaw[0]); // “number”

console.log(“Array.isArray:”, Array.isArray(queryVectorRaw)); // true

console.log(“Array.isArray:”, Array.isArray(queryVectorRaw)); // true

Has anyone experienced this before?

CodingCoop | Co-Founder Nullshot•4/22/25, 12:15 PM

Any plans to get vectorize running locally?

CCodingCoop | Co-Founder Nullshot Any plans to get vectorize running locally?

Puliczek•4/22/25, 12:48 PM

you can debug prod vectorize with "wrangler dev --experimental-vectorize-bind-to-prod"

Viktor•4/22/25, 6:59 PM

Hey everyone, is there any chance to get the amount of queries via API (see screenshot)? Already looked in the rest api, but I might be blind or there is none

PPuliczek you can debug prod vectorize with "wrangler dev --experimental-vectorize-bind-to...

CodingCoop | Co-Founder Nullshot•4/23/25, 6:57 AM

Yeah i'm familiar @Puliczek - thanks for the share - but i was on an airplane coding the other day and realized vectorize for local dev ain't a thang.

IIsaac McFadyen ?crossposting #workers-ai

Alexander•4/24/25, 9:20 AM

My bad didn't know where to post, got'ya!

#autorag found it later

nukely•4/26/25, 6:37 PM

Getting "message": "vectorize.upstream_error - failed to parse insert vectors request in ndjson format: line Some(0) was not expected format"

ac•4/28/25, 10:59 PM

Is it possible to rename a Vectorize index?

Yyevgen @beingmudit done, it should be under 30 seconds now. Thanks for reporting. Becau...

beingmudit [WePlayDOS]•4/29/25, 4:26 AM

Hey CF team, is something again wrong with Vectorize indexing? Seems like nothing gets returned even after a min. Can anyone confirm the status?

beingmudit [WePlayDOS]•4/29/25, 4:37 AM

I found an issue in my similarity pipeline that started like a day ago. The issue is now, the similarity scores have been downleveved significantly - for eg, earlier I was getting the score of 0.7 for a few content pieces and now the processing of the same content again are sending the score of ~ 0.4 hence my check is not returning results. Has this behaviour been changed recently?

nukely•5/1/25, 3:08 PM

which vector search algo is used in vectorize

Legit•5/6/25, 5:15 PM

hi does vectorize behave in a similar way to r2? in the sense that we can use preview_index_name in the wrangler.toml? just like preview_bucket_name?

thipperz•5/7/25, 4:37 PM

First time reading about precise scores when querying. How much would be the latency increase?

beingmudit [WePlayDOS]•5/8/25, 2:59 PM

Hi Team, Getting these errors very often for some time.

Vectorize API error: upstream request timeout
Can anyone verify?

Bob•5/9/25, 5:59 AM

Hello. I am building a worker application which can create vectorizedb, insert/query the vectors, delete vectorizedb. Currently, we need to bind the vectorize db to the worker manually. How can I automate it so that new dbs can be created and vectors can be inserted without manual binding in wrangler.json?

BBob Hello. I am building a worker application which can create vectorizedb, insert/q...

beingmudit [WePlayDOS]•5/9/25, 6:05 AM

Maybe you can us restful APIs if binding is an issue and i don't think dynamic bindings are supported

Bob•5/9/25, 6:08 AM

But there is a usage limit on http apis. My insert/query calls might be too many to handle in http endpoint. The only other option I see that I have to create dbs beforehand and bind them

Bbeingmudit [WePlayDOS]Maybe you can us restful APIs if binding is an issue and i don't think dynamic b...

Bob•5/9/25, 6:11 AM

thanks for the clarification

beingmudit [WePlayDOS]•5/9/25, 6:12 AM

I see

beingmudit [WePlayDOS]•5/9/25, 6:13 AM

I was not aware that bindings and rest apis have different limits

BBob But there is a usage limit on http apis. My insert/query calls might be too many...

yevgen•5/9/25, 10:51 AM

We can increase the rate limit to 9999 calls per 5 mins on your account. Please clarify what usage limits you are concerned about.

Alex Izydorczyk•5/12/25, 5:41 AM

Is the instance (index?) of vectorize creating by AutoRag feature just an ordinary instance that can be directly edited / queried

Alex Izydorczyk•5/12/25, 5:42 AM

AutoRAG today doesn’t support metadata filtering — but can I just append metadata anyway and manually query the database with these outside of the AutoRAG endpoint?

Yyevgen We can increase the rate limit to 9999 calls per 5 mins on your account. Please ...

Bob•5/14/25, 3:19 AM

insert part I can optimize but querying pattern can be unpredictable for my use case. I want to design the app with max possible throughput achievable. I am assuming that query via workers have much more throughput than http endpoints. So I will pre-create the indexes and assign them to my pipelines when needed so that I can bind the indexes in the worker before hand.

I hope there is no limit on number of vectorize bindings in a single worker unless I am reaching max worker size. I don't see any doc for that.

BBob insert part I can optimize but querying pattern can be unpredictable for my use ...

yevgen•5/14/25, 10:42 AM

you should be fine with binding many vectorize indexes in a single worker.

yevgen•5/14/25, 10:43 AM

Turned out we can only increase API rate limits for Enterprise customers, for self-serve customers we can't

ac•5/14/25, 2:31 PM

Is it still not possible to use vectorize bindings locally? I'm dumping envenv and it's completely missing, and running into errors with it. Even when I use --experimental-vectorize-bind-to-prod--experimental-vectorize-bind-to-prod it's failing (definitely have the right binding name etc)

ac•5/14/25, 2:33 PM

Weirdly, when I start up the dev server the Vectorize index shows up under [worker name] has access to the following bindings:[worker name] has access to the following bindings:. But when I try to use it in the code and when I dump env, it's not there

ac•5/14/25, 3:02 PM

Also, it works fine when running vite devvite dev but not when running

wrangler dev

wrangler dev

(I need wrangler dev because I am running two workers at once)

ac•5/14/25, 3:27 PM

Was able to get it working with the auxiliaryWorkersauxiliaryWorkers config through vite instead. Pretty broken with wrangler dev still but at least I'm unblocked. Thanks for coming to my TED talk everyone

felix_m•5/15/25, 1:05 PM

Quick question: For Vectorize billing calculation, is the total count of vectors is used? Or does namespace and metadata filtering reduced the number of billable Queried Vector DimensionsQueried Vector Dimensions for that request? I have multiple tenants with different sets of vector, so looking for an efficient way to manage these. Creating one index per tenant would work, but it seems to complicate the bindings a bit.
https://developers.cloudflare.com/vectorize/platform/pricing/

yinxingmaiming6409•5/16/25, 3:38 AM

Can't the maximum length of Vectorize be 2048?

Yyinxingmaiming6409 Can't the maximum length of Vectorize be 2048?

yevgen•5/19/25, 1:24 PM

Vectorize supports up to 1578 vector dimension

Yyevgen Vectorize supports up to 1578 vector dimension

yinxingmaiming6409•5/20/25, 3:13 AM

Can I use Vectorize using only the HTTP protocol? What is the speed limit for this access.

Yyinxingmaiming6409 Can I use Vectorize using only the HTTP protocol? What is the speed limit for th...

Viktor•5/20/25, 2:21 PM

Vectorize has a REST api -> https://developers.cloudflare.com/api/resources/vectorize/

Cloudflare API | Vectorize

Interact with Cloudflare's products and services via the Cloudflare API

Deda•5/20/25, 8:54 PM

I have a question, is it possible to make an API that takes 2 vectors from the database, and it returns a position of vector B in a table of similarity of vector A against other vectors in the database?

Deda•5/20/25, 8:54 PM

I hope the question is not confusing -.-

Deda•5/20/25, 8:56 PM

example response: Vector B is in position of 52 (meaning 52 vectors are more similar than vector B to the vector A)

Tony•5/21/25, 12:43 AM

Hi all! Quick question just about to evaluate Vectorize. Does Vectorise need a CF Worker to work? Or can I use it to connect from other clients? Thanks!

DDeda I have a question, is it possible to make an API that takes 2 vectors from the d...

Viktor•5/21/25, 9:02 AM

Not really, at least not in a clean way. You could query vector a with topK = 100 (maximum possible when not using metadata) and loop it until you find the id of vector b in there. Index 0 will be Vector A (as it's 100% equal to the one you queried), and index 99 will be the one with the least similarity. But, as said, that's not clean and only works if there are max. 98 more similar vectors then vector B

TTony Hi all! Quick question just about to evaluate Vectorize. Does Vectorise need a C...

Viktor•5/21/25, 9:02 AM

Vectorize has a REST api that does work outside workers -> https://developers.cloudflare.com/api/resources/vectorize/

Cloudflare API | Vectorize

Interact with Cloudflare's products and services via the Cloudflare API

nico escudero•5/21/25, 8:20 PM

Anyone have any repos/guides/videos of RAG pipelines being built fully on Clouldflare stack? Looking for some inspo

Dev•5/24/25, 2:46 AM

Hi, why can't we use BGE-M3 embedding model (@cf/baai/bge-m3) to create vector index?

Only these model are available to choose from: [ "@cf/baai/bge-small-en-v1.5", "@cf/baai/bge-base-en-v1.5", "@cf/baai/bge-large-en-v1.5", "openai/text-embedding-ada-002", "cohere/embed-multilingual-v2.0" ]

yinxingmaiming6409•5/24/25, 5:44 AM

When should we consider encapsulating the official SDK of C #?

DDev Hi, why can't we use BGE-M3 embedding model (@cf/baai/bge-m3) to create vector i...

yevgen•5/25/25, 11:09 AM

please ask in #workers-ai

Steak•5/25/25, 9:13 PM

Has anyone seen this Vectorize error?

VECTOR_QUERY_ERROR (code = 40006): invalid query vector, expected 1024 dimensions, and got 0 dimensions

My index is definitely set to 1024 dimensions, and I’m querying with a plain JS array of length 1024. Even a hardcoded array fails. Storage is under 600 vectors. Is this a known issue or am I missing something obvious?

Steak•5/25/25, 10:30 PM

I adjusted the query method call, and it worked.

?crossposting #workers-ai

Similar Threads

Similar Threads

Similar Threads