fal Sandbox is here - run all your models together! 🏖️
DocsBlogPricingEnterpriseCareersResearch Grants
Log-inSign-up

OpenRouter Embeddings [OpenAI Compatible] Large Language Models

openrouter/router/openai/v1/embeddings
The OpenRouter Embeddings API with fal, powered by OpenRouter, provides unified access to a wide range of large language models - including GPT, Claude, Gemini, and many others through a single API interface.
Inference
Commercial use
Schema
LLMs
PlaygroundAPI

Input

Result

Idle

Waiting for your input...

You will be charged based on the number of input and output tokens.

Logs

🧩 Usage with OpenAI Client (Embeddings API)

python
from openai import OpenAI
import os

client = OpenAI(
    base_url="https://fal.run/openrouter/router/openai/v1",
    api_key="not-needed",
    default_headers={
        "Authorization": f"Key {os.environ['FAL_KEY']}",
    },
)

response = client.embeddings.create(
    model="openai/text-embedding-3-small",
    input="An AI that learns to dream in colors humanity has never seen."
)

embedding = response.data[0].embedding
print("Embedding length:", len(embedding))

# Multiple texts example (batch encoding)
texts = [
    "An AI that learns to dream.",
    "A robot that remembers forgotten memories.",
    "A neural network that writes its own mythology.",
]

batch_response = client.embeddings.create(
    model="openai/text-embedding-3-small",
    input=texts,
)

for i, item in enumerate(batch_response.data):
    print(f"Text {i} embedding length:", len(item.embedding))

🧪 Comparing Embeddings (Simple Similarity Example)

python
from openai import OpenAI
import os
import math

client = OpenAI(
    base_url="https://fal.run/openrouter/router/openai/v1",
    api_key="not-needed",
    default_headers={
        "Authorization": f"Key {os.environ['FAL_KEY']}",
    },
)

def cosine_sim(a, b):
    dot = sum(x * y for x, y in zip(a, b))
    na = math.sqrt(sum(x * x for x in a))
    nb = math.sqrt(sum(x * x for x in b))
    return dot / (na * nb)

texts = [
    "An AI that learns to dream.",
    "A machine that hallucinates new worlds.",
    "A recipe for chocolate cake."
]

resp = client.embeddings.create(
    model="openai/text-embedding-3-small",
    input=texts,
)

emb_ai = resp.data[0].embedding
emb_machine = resp.data[1].embedding
emb_recipe = resp.data[2].embedding

print("AI vs machine:", cosine_sim(emb_ai, emb_machine))
print("AI vs recipe:", cosine_sim(emb_ai, emb_recipe))

📚 Documentation

For more details, visit the official docs: