Ask a Question

service timeout

service timeout on utc time today 00:30 am, what\`s wrong? the service department is unstable these days.

the same text but return the different vectors

we called the same text "hello" but give different result on index : 0. Is it expected? curl --location --request POST '<https://api.voyageai.com/v1/embeddings'> --header 'Authorization: Bearer --' --header 'Content-Type: application/json' --data-raw '{ "model":"voyage-large-2", "input":["hello"] }' curl --location --request POST '<https://api.voyageai.com/v1/embeddings'> --header 'Authorization: Bearer --' --header 'Content-Type: application/json' --data-raw '{ "model":"voyage-large-2", "input":["hello", "nice to meet you"] }'

Opt out of training with customer data?

Voyage AI terms of services mentioned opt-out functionality on the Website, but I can't find it on API Keys, Profile, etc. It says unless I opt out, Voyage AI could use customer data for training?

rate limit now is of 20.0 requests per minute ?not 300 per minute

errr :{"detail":"You have exceeded your Requests Per Minute (RPM) rate limit of 20.0 requests per minute. To prevent further rate limiting, try exponential backoff or sleeping between requests as described in our documentation, or try sending requests with larger batch sizes. If you would like a higher RPM}

Error: The server is overloaded or not ready yet.

Hi, We are using re rank and embbeding APIs, but we are getting a lot of errors, we had to remove from our system to normalize it, is there any status page?

Top up account

Hi I understand that you bill monthly, but I'd be good to have a top up option, to prepay for tokens before using them. I'm afraid of my bank rejecting one of your charges and having my voyageai account suspended until I speak to my bank and release payment. Thanks!

Unicode character support

We're evaluating using voyage to generate embeddings for later use in semantic search, but it appears that at least some unicode characters are not supported and cause the embedding API call to return 400 errors: `{"detail":"There was an error parsing the body"}`, for example `Pyrex Spring Blossom Light Green Beaded Edge Nested Mixing Bowl 402 1 ½ Qt` fails but `Pyrex Spring Blossom Light Green Beaded Edge Nested Mixing Bowl 402 1 Qt` succeeds. The same code framework succeeds for both inputs using gemini's text embedding API, so the basic handling of HTTP requests, content types and encodings is correct (the input is JSON encoded in utf-8, which is the assumed encoding for type application/json, and declaring it explicitly in the content-type does not change the behaviour). Does the embedding API require some transformation of the input text to restrict it to an allowed subset of unicode, or am I somehow issuing the request incorrectly?

Clustering with multilingual embeddings

Hi there Do we need to prepend "Cluster the text: " to our texts when using voyage-multilingual-2? Do you have instructions for each supported language? Thank you!

Traditional Chinese support?

Hi, does `voyage-multilingual-2`support traditional Chinese? Is there performance metric available for reference?

Where can we find benchmark results for multlingual performance on the language models?

We're trying to create a vectorstore using VoyageAI embeddings for French text. I saw one blog post vaguely mention that the rerank-1 model supports multilingual performance. Where can we find more detailed information on the multilingual performance of different models? Is there a Voyage AI embeddings model instead of a reranker that has good performance on French text?