Skip to content
Docs

AI Gateway Embeddings

Generate vector embeddings for semantic search, similarity matching, and retrieval-augmented generation (RAG).

To see which models AI Gateway supports for embeddings, use the Embedding filter at the AI Gateway Models page.

Use AI SDK 7, the AI SDK for Python beta, TanStack AI, or the OpenAI-compatible Embeddings API. Embeddings use /v1/embeddings; Chat Completions, Messages, and Responses don't accept embedding requests. Set AI_GATEWAY_API_KEY before running the examples.

Single value

For SDK options and result types, see AI SDK embeddings and Python embeddings.

app/api/embed/route.ts
import { embed } from 'ai';
 
export async function GET() {
  const result = await embed({
    model: 'openai/text-embedding-3-small',
    value: 'Sunny day at the beach',
  });
 
  return Response.json(result);
}
embed.py
import asyncio
import ai
 
async def main():
    result = await ai.ops.embed(
        ai.get_model('openai/text-embedding-3-small'),
        ["Sunny day at the beach"],
    )
    print(result.value)
 
asyncio.run(main())
embed-rest.ts
const response = await fetch('https://ai-gateway.vercel.sh/v1/embeddings', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.AI_GATEWAY_API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    model: 'openai/text-embedding-3-small',
    input: ['Sunny day at the beach'],
  }),
});
if (!response.ok) throw new Error(await response.text());
console.log(await response.json());
embed-rest.py
import json
import os
import urllib.request
 
request = urllib.request.Request(
    "https://ai-gateway.vercel.sh/v1/embeddings",
    data=json.dumps({"model": "openai/text-embedding-3-small", "input": ["Sunny day at the beach"]}).encode(),
    headers={
        'Authorization': "Bearer " + os.environ["AI_GATEWAY_API_KEY"],
        'Content-Type': "application/json"
    },
)
with urllib.request.urlopen(request) as response:
    print(json.load(response))
embed-rest.sh
curl --fail-with-body https://ai-gateway.vercel.sh/v1/embeddings \
  -H "Authorization: Bearer $AI_GATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "openai/text-embedding-3-small",
  "input": [
    "Sunny day at the beach"
  ]
}'

Multiple values

app/api/embed/route.ts
import { embedMany } from 'ai';
 
export async function GET() {
  const result = await embedMany({
    model: 'openai/text-embedding-3-small',
    values: ['Sunny day at the beach', 'Cloudy city skyline'],
  });
 
  return Response.json(result);
}
embed-many.py
import asyncio
import ai
 
async def main():
    result = await ai.ops.embed(
        ai.get_model('openai/text-embedding-3-small'),
        ["Sunny day at the beach", "Cloudy city skyline"],
    )
    print(result.value)
 
asyncio.run(main())
embed-many-rest.ts
const response = await fetch('https://ai-gateway.vercel.sh/v1/embeddings', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.AI_GATEWAY_API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    model: 'openai/text-embedding-3-small',
    input: ['Sunny day at the beach', 'Cloudy city skyline'],
  }),
});
if (!response.ok) throw new Error(await response.text());
console.log(await response.json());
embed-many-rest.py
import json
import os
import urllib.request
 
request = urllib.request.Request(
    "https://ai-gateway.vercel.sh/v1/embeddings",
    data=json.dumps({"model": "openai/text-embedding-3-small", "input": ["Sunny day at the beach", "Cloudy city skyline"]}).encode(),
    headers={
        'Authorization': "Bearer " + os.environ["AI_GATEWAY_API_KEY"],
        'Content-Type': "application/json"
    },
)
with urllib.request.urlopen(request) as response:
    print(json.load(response))
embed-many-rest.sh
curl --fail-with-body https://ai-gateway.vercel.sh/v1/embeddings \
  -H "Authorization: Bearer $AI_GATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "openai/text-embedding-3-small",
  "input": [
    "Sunny day at the beach",
    "Cloudy city skyline"
  ]
}'

Gateway provider instance

Alternatively, if you're using the Gateway provider instance, specify embedding models with gateway.textEmbeddingModel(...).

app/api/embed/route.ts
import { embed } from 'ai';
import { gateway } from '@ai-sdk/gateway';
 
export async function GET() {
  const result = await embed({
    model: gateway.textEmbeddingModel('openai/text-embedding-3-small'),
    value: 'Sunny day at the beach',
  });
 
  return Response.json(result);
}

TanStack AI

Use TanStack AI's embed function with vercelGatewayEmbedding from @tanstack/ai-vercel-gateway to generate embeddings. The adapter reads AI_GATEWAY_API_KEY, or uses VERCEL_OIDC_TOKEN when the API key is absent.

Follow the TanStack AI setup guide for installation and authentication, then use the embedding example. See the TanStack AI adapter reference for the API and result shape.

Last updated September 15, 2026
Previous
Text to Speech
Next
Reranking

Was this helpful?

supported.