Pricing
The only thing you are charged for are messages
. Retries are free.
In the pay-as-you-go plan, the price is $1 per 100K messages. Here are the price and limits of the plans.
Feature | Free | Pay as you go | Pro / Enterprise |
---|---|---|---|
Max Messages per Day | 500 | 500,000 | up to 100M |
Max Requests per Second | 100 | 100 | Custom |
Max Message Size | 1MB | 1MB | 10MB |
Max Number of Topics | 1 | 20 | infinite |
Max Number of Endpoints per Topic | 100 | 100 | infinite |
Max Retry Count | 3 | 5 | 20 |
Max Delay | 7 days | 30 days | Custom |
Max HTTP Connection Timeout | 2 min | 5 min | Custom |
Max DLQ Retention | 3 days | 7 days | 30 days |
Max DLQ Size (Number of Messages) | 1,000 | 10,000 | Custom |
Max Events Size (Number of logs to keep) | 10,000 | 10,000 | Custom |
Price per 100K Messages | N/A | $1 | Custom |
Most of the above are soft limits, contact us if you have custom needs.
Free tier
The free tier includes 500 Messages per day. Afterwards we will no longer accept new requests.
Pay as You Go
Upstash limits your monthly usage for QStash to daily 500,000 in pay-as-you-go plan. If you are expecting to exceed this limit, contact us for the Pro or Enterprise plans.
Pro / Enterprise
We give you a fixed price for higher throughput needs and custom quotas. We reserve isolated resource to guarantee the scalability for high volumes. Here are the plans and prices:
Plan | Pro 1M | Pro 10M | Enterprise |
---|---|---|---|
Max Messages / Day | 1M | 10M | 100M |
Max Messages / Sec | 1000 | 1000 | Custom |
Price / Month | $180 | $420 | Contact Us |
Pro and Enterprise plans have a minimum term of 1 month, afterwards you can cancel any time.
Examples:
Publish to single API endpoint
- You make a request to
/v2/publish/<your-api-url>
- We make an HTTP request to your API and it returns a 500 status code
- After some time we retry to deliver the message and it succeeds
In total, you would be charged for 1 message.
Example “every hour”
When your schedule triggers every hour, at the end of the month this would be
around 24 * 30 = 720
billed messages.
Publish to topic
Let’s assume you have a topic and 2 endpoints (A
and B
) subscribed to it.
- You make a request to
/v2/publish/<your-topic>
- We make an HTTP request to each of your endpoints.
In total this would be 2 messages.