GPT-5 Mini API Pricing: $0.25/$2.00 Per Million Tokens

Q: How much does GPT-5 mini cost?

GPT-5 mini costs $0.25 per million input tokens and $2.00 per million output tokens. This is 70% cheaper than GPT-5.4 ($2.50/$15.00) while offering strong quality for most production workloads.

Q: What is the context window of GPT-5 mini?

GPT-5 mini has a 272K token context window. This is larger than GPT-5 nano's 128K and sufficient for most document processing, chat, and summarization tasks.

Q: What is GPT-5 mini good for?

GPT-5 mini excels at chatbots, content moderation, summarization, Q&A systems, and classification. It offers a strong balance of quality and cost — better reasoning than GPT-5 nano at 5x the price, and 70% cheaper than GPT-5.4 for tasks that don't need top-tier reasoning.

GPT-5 mini is 70% cheaper than GPT-5.4 with strong reasoning quality — but shutting down Dec 11, 2026. Here's what you need to know.

Updated Aug 14, 2026 · 94 models tracked across 11 providers

TL;DR

Price: $0.25/M input, $2.00/M output — 70% cheaper than GPT-5.4 ($2.50/$15.00)
Context: 272K tokens — handles large documents and multi-turn conversations
Provider: OpenAI API (platform.openai.com)
Best for: Chatbots, content moderation, summarization, Q&A, classification
Deprecation: Shutting down Dec 11, 2026 (~4 months left). Migrate to GPT-5.6 Terra ($2.00/$12.00) or GPT-5.6 Luna ($0.20/$1.20)
Trade-off: 5x more expensive than GPT-5 nano ($0.05/$0.40) but significantly better reasoning

Deprecation Notice: GPT-5 Mini Shutting Down Dec 11, 2026

OpenAI has announced that GPT-5 mini will be retired on December 11, 2026. After this date, API calls to GPT-5 mini will fail. You have approximately 4 months to migrate.

Migration options:

GPT-5.6 Terra ($2.00/$12.00) — the official replacement. Higher quality than GPT-5 mini, closer to GPT-5.4. Best if you need the strongest reasoning in this price range.
GPT-5.6 Luna ($0.20/$1.20) — a budget alternative that's actually cheaper than GPT-5 mini. Good option if cost is the priority and you can accept slightly lower quality than Terra.
GPT-5 nano ($0.05/$0.40) — for ultra-cheap, high-volume tasks like classification and extraction. Limited reasoning but unbeatable on cost.

We recommend starting migration now. Switch to GPT-5.6 Luna first — it's cheaper and likely covers most GPT-5 mini use cases. Upgrade to Terra only if quality tests show a regression.

GPT-5 Mini Pricing Breakdown

At $0.25 per million input tokens and $2.00 per million output tokens, GPT-5 mini is 70% cheaper than GPT-5.4 ($2.50/$15.00) on both input and output. Here's how the costs scale:

Monthly Volume	Input Cost	Output Cost	Total Cost
1M tokens	$0.25	$2.00	$2.25
10M tokens	$2.50	$20.00	$22.50
100M tokens	$25.00	$200.00	$225.00
1B tokens	$250.00	$2,000.00	$2,250.00

At 100M tokens/month, GPT-5 mini costs $225. The same volume on GPT-5.4 would cost $1,275 — a $1,050/month savings. On GPT-5 nano it would be just $45, but with weaker reasoning.

How GPT-5 Mini Compares to Other Budget Models

Model	Input $/M	Output $/M	Context	Provider
GPT-5 nano	$0.05	$0.40	128K	OpenAI
GPT-5.4 nano	$0.20	$1.25	128K	OpenAI
Qwen 3.7 Flash	$0.03	$0.13	1M	Alibaba
GPT-5 mini	$0.25	$2.00	272K	OpenAI
Gemini 2.5 Flash-Lite	$0.10	$0.40	1M	Google
Ministral 3 3B	$0.10	$0.10	128K	Mistral

GPT-5 mini sits in a unique position: it's significantly cheaper than GPT-5.4 while offering much better reasoning than nano-class models. The 272K context window is larger than most OpenAI budget models (128K) but smaller than Qwen 3.7 Flash or Gemini 2.5 Flash-Lite (1M).

GPT-5 Mini vs GPT-5 Nano: Is 5x Worth It?

GPT-5 mini costs 5x more than GPT-5 nano on input and output. Here's what that premium gets you:

Feature	GPT-5 nano	GPT-5 mini
Input price	$0.05/M	$0.25/M
Output price	$0.40/M	$2.00/M
Context window	128K	272K
Reasoning	Basic	Strong
Function calling	Yes	Yes
Structured outputs	Yes	Yes
Best for	Classification, extraction	Chatbots, summarization, Q&A

Rule of thumb: Use GPT-5 nano for simple classification, extraction, and moderation tasks. Upgrade to GPT-5 mini when you need better reasoning, longer context, or more nuanced responses — like customer-facing chatbots or document summarization.

When to Use GPT-5 Mini

Chatbots

Customer support, sales assistants, and conversational agents. Strong enough for multi-turn dialogue at a fraction of GPT-5.4's cost.

Content Moderation

Flag policy violations, spam, and inappropriate content at scale. Better judgment than nano models for edge cases.

Summarization

Condense documents, meeting notes, and support tickets. 272K context handles long documents in a single pass.

Q&A Systems

Knowledge base lookups, FAQ bots, and internal helpdesks. Provides accurate, well-structured answers from context.

Classification

Intent detection, sentiment analysis, ticket routing. More accurate than nano models on ambiguous inputs.

Draft Generation

Email drafts, report outlines, first-pass content. Good quality at budget pricing for non-critical output.

When NOT to Use GPT-5 Mini

GPT-5 mini is a strong mid-tier model, but it has limits. Upgrade or downgrade when:

Complex reasoning: Multi-step logic, advanced math, or complex code generation → use GPT-5.4 ($2.50/$15) or Claude Sonnet 5 ($2/$10)
Critical content: Customer-facing marketing, legal documents, or creative writing → use Claude Opus 5 ($5/$25)
Ultra-high volume: Millions of simple classifications or extractions → save money with GPT-5 nano ($0.05/$0.40)
Very large context: Documents over 272K tokens → use Qwen 3.7 Flash (1M, $0.03/$0.13) or Gemini 2.5 Flash-Lite (1M, $0.10/$0.40)
Long-term projects: GPT-5 mini shuts down Dec 11, 2026 → start migrating to GPT-5.6 Luna ($0.20/$1.20) or Terra ($2.00/$12.00) now

Real-World Cost Scenario

Let's say you're running 50,000 customer support conversations per month, with an average of 1,500 input tokens and 500 output tokens per conversation:

Model	Monthly Input	Monthly Output	Total
GPT-5 nano	$3.75	$10.00	$13.75
GPT-5.4 nano	$15.00	$31.25	$46.25
GPT-5 mini	$18.75	$50.00	$68.75
Gemini 2.5 Flash-Lite	$7.50	$10.00	$17.50
GPT-5.4	$187.50	$375.00	$562.50

GPT-5 mini costs $68.75/month for 50K support conversations — 8x cheaper than GPT-5.4 ($562.50) while delivering strong conversational quality. Gemini 2.5 Flash-Lite is cheaper at $17.50, but GPT-5 mini offers better reasoning and OpenAI ecosystem integration.

Frequently Asked Questions

How much does GPT-5 mini cost?

$0.25 per million input tokens and $2.00 per million output tokens. This is 70% cheaper than GPT-5.4 ($2.50/$15.00) while offering strong quality for most production workloads.

Is GPT-5 mini being deprecated?

Yes. GPT-5 mini will be shut down on December 11, 2026. It will be replaced by GPT-5.6 Terra ($2.00/$12.00). For a budget alternative, consider GPT-5.6 Luna ($0.20/$1.20) or GPT-5 nano ($0.05/$0.40).

What is the context window of GPT-5 mini?

272K tokens. This is larger than GPT-5 nano's 128K and sufficient for most document processing, chat, and summarization tasks.

Should I migrate from GPT-5 mini to GPT-5.6 Luna or GPT-5.6 Terra?

If you need quality close to GPT-5.4, migrate to GPT-5.6 Terra ($2.00/$12.00). If budget is the priority, GPT-5.6 Luna ($0.20/$1.20) is actually cheaper than GPT-5 mini with comparable quality. For ultra-cheap tasks, GPT-5 nano ($0.05/$0.40) remains the lowest-cost option.

What is GPT-5 mini good for?

Chatbots, content moderation, summarization, Q&A systems, and classification. It offers a strong balance of quality and cost — better reasoning than GPT-5 nano at 5x the price, and 70% cheaper than GPT-5.4 for tasks that don't need top-tier reasoning.

Calculate Your GPT-5 Mini Costs

Enter your token counts and see exactly what you'd pay. Compare against 93 other models instantly.

Open Calculator → See Full Ranking