Discussions
Accessing Billing and Usage Information via API
Hello,
Applying Semantic Search to the Friday Night Funkin' Mod Ecosystem using Voyage AI Embeddings
Hello Voyage AI Community,
As a developer and a fan of the rhythm game Friday Night Funkin' (FNF), I’ve been exploring ways to improve how fans discover new content. With the FNF modding community producing thousands of unique mods (on platforms like GameBanana), finding a specific mod based on "vibe," "music style," or "difficulty level" is becoming increasingly difficult with standard keyword search.
I am considering a project that uses Voyage AI’s embeddings (like voyage-large-2) to build a semantic search engine for FNF mods. Here is the concept:
The Challenge: FNF mods have rich metadata—descriptions of the music (jazz, metal, lo-fi), the character's backstory, and mechanical difficulty. A simple search for "fast songs" might miss mods described as "high BPM" or "intense speed."
The Solution with Voyage AI: By embedding the descriptions and tags of thousands of FNF mods, we can create a vector space where users can search using natural language (e.g., "A mod with a creepy atmosphere and challenging electronic music").
Potential Implementation: Using Voyage AI's high-quality rerankers to ensure that the most relevant mods appear at the top, even if the user's query doesn't match the exact title.
Why FNF?
FNF is open-source and has one of the most active creative communities. It’s a perfect "stress test" for embedding models because the language used in the community is very informal and niche-specific.
I would love to hear your thoughts on:
Which Voyage model would be best for handling gaming-related slang and technical music terms?
Has anyone tried building a recommendation engine for gaming assets using Voyage AI yet?
Looking forward to your insights!
Deprecation of models
Does VoyageAI have a policy about how much advance notice will be provided when a model is going to be deprecated?
Is the 200M free tokens shared or per model?
The docs say we get 200 million free tokens per account for embeddings. Is that quota shared across all models (like voyage-3.5, voyage-3-large, etc.) or does each model get its own 200M?
OpenAI API compatibitlity
To be compatible with OpenAI API implementation Voyage AI API should also deliver "prompt_tokens" in the "usage" part of the response for Embeddings.
Language Support in Embedding model
Hi there, can you please list the languages that Voyage AI's embedding model natively supports?
How to get 2048 byte embedding for voyage-3.5 model?
Hi, I'm using Typescript library to get voyage-3.5 embedding. Despite the documentation contains description of output_dimension parameetr of a request, Typescript library does not support it. method embed() takes a VoyageAI.EmbedRequest type object which contains only input, model, inputType, truncation, and encodingFormat properties. There is neither output_dimension, nor output_dtype. Even if I try to create a request of any type and include output_dimension: 2048 in it, the embed method still returns 1024 embedding vector. How can I request a different embedding size?
too many concurrent streams
Running into too many concurrent streams exception -- I take that is a bit different than requests per minute or maximum tokens being exceeded. How many concurrent streams are allowed?
Is Voyage multimodal embedding api supports sinhala language?
I'm on a big, high accuracy needed project. so I want to know that voyage multimodal embedding api will do it's job with best performance for sinhala language
Is rerank-3 coming soon?
Lots of new stuff in the 3rd generation of embeddings like voyage-3-large, 3.5 series and contextual embeddings.