OpenAI frontier models
GPT-5 Mini pricing
Ideal for Tier-1 support, notification digests, and experimentation before upgrading to GPT-5.x premium tiers.
Last pricing check: Mar 13, 2026
Why teams choose this model
Scenario planning
Realistic cost examples
Numbers use GPT-5 Mini pricing
Tier-1 chat
Support teams offload routine cases to GPT-5 Mini with cached instructions.
Per request
$0.0007
Per month
$87.00
Tokens sent
96,000,000
Notification digest
Ops digests summarize alerts for Slack/email at pennies per request.
Per request
$0.0006
Per month
$54.00
Tokens sent
58,500,000
Experimentation sandbox
Product squads prototype flows cheaply before graduating to larger GPT-5 siblings.
Per request
$0.001
Per month
$67.38
Tokens sent
73,500,000
Compare with
FAQs
How much do cache hits save?
When prompts share long prefixes, cached input drops from $0.25 to $0.025 per million tokens—a 10x savings.
Can GPT-5 Mini handle tools?
Yes, it supports the same Responses/Assistants APIs. TokenTally’s pricing matches the official table so you can plan tool-heavy workloads.
What if I outgrow Mini?
Jump to GPT-5.1/5.2 for better reasoning while keeping costs manageable; this guide helps you see the step-change in spend.
Pricing sources
- https://developers.openai.com/api/docs/pricing
Checked Mar 13, 2026