Skip to main content
Fhddos’s embeddings endpoint converts text into high-dimensional vectors using Google’s Gemini embedding models. These vectors capture semantic meaning, so similar texts produce similar vectors — making them the foundation for semantic search, retrieval-augmented generation (RAG), recommendation systems, and text classification. Authenticate with your Fhddos token; no Google API key is needed.

​
Set Up Your Environment

export BASE_URL="https://aiapi.fhddos.com"
export TOKEN="your-fhddos-token"

​
Generate Embeddings: POST /v1/embeddings

​
Single Input

curl -X POST "$BASE_URL/v1/embeddings" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text-embedding-004",
    "input": "The quick brown fox jumps over the lazy dog"
  }'

​
Response

{
  "object": "list",
  "data": [
    {
      "embedding": [0.0123, -0.0456, 0.0789, "..."],
      "index": 0,
      "object": "embedding"
    }
  ],
  "model": "text-embedding-004",
  "usage": {
    "prompt_tokens": 9,
    "total_tokens": 9
  }
}
The embedding array contains the full vector. For text-embedding-004 and gemini-embedding-001, this is a 768-dimensional float array.

​
Batch Input

Pass an array of strings to embed multiple texts in a single request: 
curl -X POST "$BASE_URL/v1/embeddings" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text-embedding-004",
    "input": [
      "How do I reset my password?",
      "Where can I find my invoice?",
      "What are your support hours?"
    ]
  }'
The response data array contains one embedding object per input string, each with a matching index.

​
Available Models

ModelDimensionsDescription
text-embedding-004768General-purpose text embedding — recommended starting point
gemini-embedding-001768Gemini native embedding model

​
SDK Examples

from openai import OpenAI

client = OpenAI(
    api_key="your-fhddos-token",
    base_url="https://aiapi.fhddos.com/v1"
)

response = client.embeddings.create(
    model="text-embedding-004",
    input="The quick brown fox jumps over the lazy dog"
)

vector = response.data[0].embedding
print(f"Vector dimensions: {len(vector)}")
print(f"First 5 values: {vector[:5]}")

​
Common Use Cases

Semantic Search

Embed your document corpus and a user’s query, then rank documents by cosine similarity to surface the most relevant results — without keyword matching.

Retrieval-Augmented Generation (RAG)

Store chunk embeddings in a vector database (Pinecone, pgvector, Weaviate). At query time, retrieve the top-k nearest chunks and pass them as context to a Gemini chat model.

Text Clustering & Deduplication

Group similar documents together or detect near-duplicate content by comparing embedding distances across your dataset.

Recommendation Systems

Compute similarity between item descriptions or user preference vectors to power content or product recommendations.

​
Measuring Similarity

Use cosine similarity to compare two embedding vectors. Values close to 1.0 indicate high semantic similarity; values near 0 indicate unrelated content. 
import numpy as np
from openai import OpenAI

client = OpenAI(
    api_key="your-fhddos-token",
    base_url="https://aiapi.fhddos.com/v1"
)

def embed(text: str) -> list[float]:
    resp = client.embeddings.create(model="text-embedding-004", input=text)
    return resp.data[0].embedding

def cosine_similarity(a: list[float], b: list[float]) -> float:
    a, b = np.array(a), np.array(b)
    return float(np.dot(a, b) / (np.linalg.norm(a) * np.linalg.norm(b)))

vec_a = embed("How do I cancel my subscription?")
vec_b = embed("Steps to unsubscribe from the service")
vec_c = embed("What is the capital of France?")

print(cosine_similarity(vec_a, vec_b))  # High — semantically similar
print(cosine_similarity(vec_a, vec_c))  # Low  — unrelated topics