GPT-5.4 Mini vs GPT-5.4 Nano

GPT-5.4 Nano undercuts GPT-5.4 Mini right across the rate card: $0.20 against $0.75 per 1M for input, $1.25 against $4.50 for output, 73% and 72% less. A single chat turn costs a fraction of a cent on both, so this only becomes real money at volume: a thousand long documents costs $83.75 against $313.50.

GPT-5.4 Mini costs 3.6 times as much as GPT-5.4 Nano, and what that buys is 3.4 points of overall intelligence (Intelligence Index): 24.1 (xhigh) against 20.7 (xhigh). Unless you genuinely need that lead, take GPT-5.4 Nano: GPT-5.4 Mini is 3.6 times the price.

Standard pay-as-you-go prices. Last updated 27 Sep 2026.

Never miss an AI pricing or launch update

Get notified about new model launches, pricing changes, provider updates, deprecations, and cost-saving insights.

๐Ÿ”’ We respect your privacy. Unsubscribe anytime.

Cheaper overall

GPT-5.4 Nano

73% lower blended rate

Higher benchmark scores

GPT-5.4 Mini

leads on 10 of 13 figures

Cheaper cached input

GPT-5.4 Nano

$0.02 vs $0.075, 73% less

Where each one wins

GPT-5.4 Mini

  • Higher overall intelligence (Intelligence Index) 24.1 vs 20.7
  • Higher multi-step tool use (Agentic Index) 17.9 vs 16
  • Higher graduate-level science (GPQA Diamond) 87.5 vs 81.7
  • Ahead on 7 more benchmark figures

GPT-5.4 Nano

  • Cheaper input tokens $0.20 vs $0.75, 73% less
  • Cheaper output tokens $1.25 vs $4.50, 72% less
  • Cheaper cached input $0.02 vs $0.075, 73% less
  • Cheaper on all four workloads
  • Higher expert-exam performance (Humanity's Last Exam) 28.3 vs 28.1
  • Higher instruction following (IFBench) 75.9 vs 73.3
  • Higher answer reliability (Omniscience Non-Hallucination) 48.9 vs 10.2

What four workloads cost

One run, and the same run a thousand times.

GPT-5.4 Nano is cheaper on all four workloads, by much the same margin each time โ€” about 3.7 times, from $0.0008 against $0.0030 on a chat turn to $0.0838 against $0.3135 on a long document.

Workload GPT-5.4 Mini GPT-5.4 Nano ร—1,000 Difference
Chat turn 1,000 in / 500 out $0.0030 $0.0008 $3.00 / $0.8250 GPT-5.4 Nano is 73% cheaper
RAG answer 10,000 in / 800 out $0.0111 $0.0030 $11.10 / $3.00 GPT-5.4 Nano is 73% cheaper
Agent loop 32,000 in / 700 out ร— 20 calls $0.1380 $0.0375 $138.00 / $37.50 GPT-5.4 Nano is 73% cheaper
Long document 400,000 in / 3,000 out $0.3135 $0.0838 $313.50 / $83.75 GPT-5.4 Nano is 73% cheaper

Price your own token counts for these two โ†’

Price per 1M tokens

Above 272,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point. It does not change which of the two is cheaper.

Context band Price GPT-5.4 Mini GPT-5.4 Nano Difference
Prompts up to 272,000 tokens Input $0.75 $0.20 GPT-5.4 Nano is 73% cheaper
Prompts up to 272,000 tokens Cached input $0.075 $0.02 GPT-5.4 Nano is 73% cheaper
Prompts up to 272,000 tokens Output $4.50 $1.25 GPT-5.4 Nano is 72% cheaper
Prompts over 272,000 tokens Input $0.75 $0.20 GPT-5.4 Nano is 73% cheaper
Prompts over 272,000 tokens Cached input $0.075 $0.02 GPT-5.4 Nano is 73% cheaper
Prompts over 272,000 tokens Output $4.50 $1.25 GPT-5.4 Nano is 72% cheaper

Blended: GPT-5.4 Mini $1.6875, GPT-5.4 Nano $0.4625 per 1M โ€” one rate at 3:1 input to output, for comparing two models at a glance.

Both priced by OpenAI. Standard tier, pay-as-you-go.

Other pricing tiers

Both models offer Batch and Flex, worth up to 50% off their standard rates, and the cheaper model is the same one on every tier โ€” so the tier you buy changes the bill rather than the choice.

Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate โ€” how we read tiers.

Tier GPT-5.4 Mini GPT-5.4 Nano Cheaper
Batch $0.375 in / $2.25 out 50% off Standard $0.10 in / $0.625 out 50% off Standard GPT-5.4 Nano
Flex $0.375 in / $2.25 out 50% off Standard $0.10 in / $0.625 out 50% off Standard GPT-5.4 Nano

Benchmarks

They share 14 figures: GPT-5.4 Mini leads on 10, GPT-5.4 Nano on 3. The widest gap is 38.7 points, on answer reliability (Omniscience Non-Hallucination).

Overall intelligence

Intelligence Index โ€” Overall intelligence across reasoning, knowledge & math evals

GPT-5.4 Mini 24.1
GPT-5.4 Nano 20.7

GPT-5.4 Mini ahead by 3.4

Coding ability

Coding Index โ€” Coding ability across software-engineering evals

GPT-5.4 Mini 56.1
GPT-5.4 Nano 56.1

Level

Multi-step tool use

Agentic Index โ€” Tool use & multi-step agent task performance

GPT-5.4 Mini 17.9
GPT-5.4 Nano 16

GPT-5.4 Mini ahead by 1.9

Graduate-level science

GPQA Diamond โ€” Graduate-level questions in biology, chemistry & physics

GPT-5.4 Mini 87.5
GPT-5.4 Nano 81.7

GPT-5.4 Mini ahead by 5.8

Expert-exam performance

Humanity's Last Exam โ€” Expert-level questions across many academic domains

GPT-5.4 Mini 28.1
GPT-5.4 Nano 28.3

GPT-5.4 Nano ahead by 0.2

Instruction following

IFBench โ€” Precise following of detailed instructions

GPT-5.4 Mini 73.3
GPT-5.4 Nano 75.9

GPT-5.4 Nano ahead by 2.6

Support-agent performance

ฯ„ยฒ-Bench Telecom โ€” Tool-using agent tasks in a telecom support setting

GPT-5.4 Mini 83.3
GPT-5.4 Nano 76

GPT-5.4 Mini ahead by 7.3

Long-context reasoning

AA-LCR โ€” Long-context reasoning across large inputs

GPT-5.4 Mini 77
GPT-5.4 Nano 76.7

GPT-5.4 Mini ahead by 0.3

Scientific coding

SciCode โ€” Research-level scientific coding tasks

GPT-5.4 Mini 52.1
GPT-5.4 Nano 47.2

GPT-5.4 Mini ahead by 4.9

Command-line work

Terminal-Bench Hard โ€” Complex command-line and terminal workflows

GPT-5.4 Mini 52.3
GPT-5.4 Nano 42.4

GPT-5.4 Mini ahead by 9.9

Physics research reasoning

CritPt โ€” Unpublished physics research reasoning problems

GPT-5.4 Mini 10
GPT-5.4 Nano 9.3

GPT-5.4 Mini ahead by 0.7

Real-world task performance

GDPval โ€” Economically valuable, real-world knowledge work

GPT-5.4 Mini 25
GPT-5.4 Nano 21.8

GPT-5.4 Mini ahead by 3.2

Factual accuracy

Omniscience Accuracy โ€” Breadth of factual knowledge across domains

GPT-5.4 Mini 37.5
GPT-5.4 Nano 25.7

GPT-5.4 Mini ahead by 11.8

Answer reliability

Omniscience Non-Hallucination โ€” How reliably the model avoids fabricated answers

GPT-5.4 Mini 10.2
GPT-5.4 Nano 48.9

GPT-5.4 Nano ahead by 38.7

Each figure is that model's best published run, with the effort level named beside it โ€” where these scores come from.

Specifications

Specification GPT-5.4 Mini GPT-5.4 Nano
Context window 400,000 tokens 400,000 tokens
Max output 128,000 tokens 128,000 tokens
Takes in Text, image, file Text, image, file
Puts out Text Text
Knowledge cutoff 31 Aug 2025 31 Aug 2025
Released 17 Mar 2026 17 Mar 2026
Status Active Active
Sold by OpenAI, Perplexity OpenAI, Perplexity

FAQs

Is GPT-5.4 Mini cheaper than GPT-5.4 Nano?
No โ€” the other way round. A 1,000-token prompt with a 500-token reply costs $0.0030 on GPT-5.4 Mini and $0.0008 on GPT-5.4 Nano, and the same model is cheaper on every workload on this page.
How much do GPT-5.4 Mini and GPT-5.4 Nano cost per 1M tokens?
GPT-5.4 Mini costs $0.75 for input and $4.50 for output. GPT-5.4 Nano costs $0.20 and $1.25. Standard pay-as-you-go rates.
Which is better, GPT-5.4 Mini or GPT-5.4 Nano?
On published benchmarks, GPT-5.4 Mini. It leads on 10 of the 14 figures both models report, including overall intelligence (Intelligence Index), where it scores 24.1 against 20.7.
Is Batch pricing cheaper for GPT-5.4 Mini or GPT-5.4 Nano?
Batch input costs $0.375 on GPT-5.4 Mini and $0.10 on GPT-5.4 Nano, 50% below the standard rate on both.

Explore these models

GPT-5.4 Mini

OpenAI ยท 400,000-token context

Pricing · Calculator · Specifications

GPT-5.4 Nano

OpenAI ยท 400,000-token context

Pricing · Calculator · Specifications

Other comparisons