Service Tiers
Last updated March 18, 2026
OpenAI offers different processing tiers that trade off between latency, availability, and cost. You can pass the service_tier parameter through AI Gateway to control which tier OpenAI uses for your request. AI Gateway automatically adjusts pricing based on the tier used.
Service tiers are currently only supported for OpenAI models routed through the OpenAI provider. If you set service_tier for a non-OpenAI model, the parameter is ignored.
| Value | Description |
|---|---|
default | Standard processing tier |
priority | Higher availability and faster processing at increased cost |
flex | Lower cost with potentially higher latency |
If you don't specify service_tier, requests use the standard tier by default.
app/api/chat/route.ts
import { generateText } from 'ai';
const { text, usage, providerMetadata } = await generateText({
model: 'openai/gpt-5',
prompt: 'Explain quantum computing in two sentences.',
providerOptions: {
openai: {
serviceTier: 'flex',
},
},
});
console.log(text);
console.log('Service tier:', providerMetadata?.openai?.serviceTier);
console.log('Usage:', usage);app/api/chat/route.ts
import { gateway } from '@ai-sdk/gateway';
import { generateText } from 'ai';
const { text, usage, providerMetadata } = await generateText({
model: gateway('openai/gpt-5'),
prompt: 'Explain quantum computing in two sentences.',
providerOptions: {
openai: {
serviceTier: 'flex',
},
},
});
console.log(text);
console.log('Service tier:', providerMetadata?.openai?.serviceTier);
console.log('Usage:', usage);service-tier.ts
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.AI_GATEWAY_API_KEY,
baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const response = await client.chat.completions.create({
model: 'openai/gpt-5',
messages: [
{
role: 'user',
content: 'Explain quantum computing in two sentences.',
},
],
service_tier: 'flex',
});
console.log(response.choices[0].message.content);
console.log('Service tier:', response.service_tier);
console.log('Usage:', response.usage);service-tier.py
import os
from openai import OpenAI
client = OpenAI(
api_key=os.getenv('AI_GATEWAY_API_KEY'),
base_url='https://ai-gateway.vercel.sh/v1'
)
response = client.chat.completions.create(
model='openai/gpt-5',
messages=[
{
'role': 'user',
'content': 'Explain quantum computing in two sentences.'
}
],
service_tier='flex'
)
print(response.choices[0].message.content)
print('Service tier:', response.service_tier)
print('Usage:', response.usage)service-tier.ts
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.AI_GATEWAY_API_KEY,
baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const response = await client.responses.create({
model: 'openai/gpt-5',
input: 'Explain quantum computing in two sentences.',
service_tier: 'flex',
});
console.log(response.output_text);
console.log('Service tier:', response.service_tier);
console.log('Usage:', response.usage);service-tier.py
import os
from openai import OpenAI
client = OpenAI(
api_key=os.getenv('AI_GATEWAY_API_KEY'),
base_url='https://ai-gateway.vercel.sh/v1'
)
response = client.responses.create(
model='openai/gpt-5',
input='Explain quantum computing in two sentences.',
service_tier='flex'
)
print(response.output_text)
print('Service tier:', response.service_tier)
print('Usage:', response.usage)Service tiers work with streaming requests. The service_tier field appears in the response:
app/api/chat/route.ts
import { streamText } from 'ai';
const result = streamText({
model: 'openai/gpt-5',
prompt: 'Explain quantum computing in two sentences.',
providerOptions: {
openai: {
serviceTier: 'priority',
},
},
});
for await (const textPart of result.textStream) {
process.stdout.write(textPart);
}
const { usage, providerMetadata } = await result;
console.log('Service tier:', providerMetadata?.openai?.serviceTier);
console.log('Usage:', usage);app/api/chat/route.ts
import { gateway } from '@ai-sdk/gateway';
import { streamText } from 'ai';
const result = streamText({
model: gateway('openai/gpt-5'),
prompt: 'Explain quantum computing in two sentences.',
providerOptions: {
openai: {
serviceTier: 'priority',
},
},
});
for await (const textPart of result.textStream) {
process.stdout.write(textPart);
}
const { usage, providerMetadata } = await result;
console.log('Service tier:', providerMetadata?.openai?.serviceTier);
console.log('Usage:', usage);service-tier-streaming.ts
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.AI_GATEWAY_API_KEY,
baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const stream = await client.chat.completions.create({
model: 'openai/gpt-5',
messages: [
{
role: 'user',
content: 'Explain quantum computing in two sentences.',
},
],
stream: true,
service_tier: 'priority',
});
for await (const chunk of stream) {
const content = chunk.choices[0]?.delta?.content;
if (content) process.stdout.write(content);
}service-tier-streaming.py
import os
from openai import OpenAI
client = OpenAI(
api_key=os.getenv('AI_GATEWAY_API_KEY'),
base_url='https://ai-gateway.vercel.sh/v1'
)
stream = client.chat.completions.create(
model='openai/gpt-5',
messages=[
{
'role': 'user',
'content': 'Explain quantum computing in two sentences.'
}
],
stream=True,
service_tier='priority'
)
for chunk in stream:
content = chunk.choices[0].delta.content
if content:
print(content, end='', flush=True)service-tier-streaming.ts
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.AI_GATEWAY_API_KEY,
baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const stream = await client.responses.create({
model: 'openai/gpt-5',
input: 'Explain quantum computing in two sentences.',
stream: true,
service_tier: 'priority',
});
for await (const event of stream) {
if (event.type === 'response.output_text.delta') {
process.stdout.write(event.delta);
}
}service-tier-streaming.py
import os
from openai import OpenAI
client = OpenAI(
api_key=os.getenv('AI_GATEWAY_API_KEY'),
base_url='https://ai-gateway.vercel.sh/v1'
)
stream = client.responses.create(
model='openai/gpt-5',
input='Explain quantum computing in two sentences.',
stream=True,
service_tier='priority'
)
for event in stream:
if event.type == 'response.output_text.delta':
print(event.delta, end='', flush=True)AI Gateway adjusts pricing based on the service tier used. The tables below show per-million-token rates.
For the most up-to-date pricing, refer to the OpenAI pricing page.
| Model | Input | Output | Cached input |
|---|---|---|---|
| gpt-5.4 | $5.00 | $30.00 | $0.50 |
| gpt-5.2 | $3.50 | $28.00 | $0.35 |
| gpt-5.1 | $2.50 | $20.00 | $0.25 |
| gpt-5 | $2.50 | $20.00 | $0.25 |
| gpt-5-mini | $0.45 | $3.60 | $0.045 |
| gpt-5.3-codex | $3.50 | $28.00 | $0.35 |
| gpt-5.2-codex | $3.50 | $28.00 | $0.35 |
| gpt-5.1-codex-max | $2.50 | $20.00 | $0.25 |
| gpt-5.1-codex | $2.50 | $20.00 | $0.25 |
| gpt-5-codex | $2.50 | $20.00 | $0.25 |
| gpt-4.1 | $3.50 | $14.00 | $0.875 |
| gpt-4.1-mini | $0.70 | $2.80 | $0.175 |
| gpt-4.1-nano | $0.20 | $0.80 | $0.05 |
| gpt-4o | $4.25 | $17.00 | $2.125 |
| gpt-4o-2024-05-13 | $8.75 | $26.25 | — |
| gpt-4o-mini | $0.25 | $1.00 | $0.125 |
| o3 | $3.50 | $14.00 | $0.875 |
| o4-mini | $2.00 | $8.00 | $0.50 |
| Model | Input | Output | Cached input |
|---|---|---|---|
| gpt-5.4 | $1.25 | $7.50 | $0.13 |
| gpt-5.4-pro | $15.00 | $90.00 | — |
| gpt-5.4-mini | $0.375 | $2.25 | $0.0375 |
| gpt-5.4-nano | $0.10 | $0.625 | $0.01 |
| gpt-5.2 | $0.875 | $7.00 | $0.0875 |
| gpt-5.1 | $0.625 | $5.00 | $0.0625 |
| gpt-5 | $0.625 | $5.00 | $0.0625 |
| gpt-5-mini | $0.125 | $1.00 | $0.0125 |
| gpt-5-nano | $0.025 | $0.20 | $0.0025 |
| o3 | $1.00 | $4.00 | $0.25 |
| o4-mini | $0.55 | $2.20 | $0.138 |
Was this helpful?