OpenAIGPT-5 miniVSOpenAIGPT-5.4 mini
Our Take
We recommend GPT-5.4 mini for its clear benchmark advantage, or the 2.5x cheaper GPT-5 mini only if your budget requires optimizing costs for very high-volume pipelines. While GPT-5.4 mini offers superior reasoning and coding, it carries a moderate price premium. Choose GPT-5.4 mini for quality, or GPT-5 mini for cost optimization.
▶WHY?
Benchmark Calculations & Evidence:
Coding Benchmarks: Both models were evaluated on the SWE-bench Pro benchmark. GPT-5.4 mini scored 54.4%, while GPT-5 mini scored 45.7%.
Reasoning Benchmarks: Both models were evaluated on the GPQA Diamond benchmark. GPT-5.4 mini scored 87.5%, while GPT-5 mini scored 81.6%.
Cost Efficiency: GPT-5 mini pricing ($0.25/M input, $2/M output) is 2.5x cheaper than GPT-5.4 mini ($0.75/M input, $4.5/M output).
Was this recommendation helpful?
Benchmarks & Scores
Coding (swe-bench-pro)
45.7%complex codebases, multi-file repositories, and architectural planning
Reasoning (gpqa-diamond)
81.6%graduate-level science QA
Cost & Context
Cost (per 1M tokens)2.5x cheaper
$0.69Input: $0.25 | Output: $2.00Context Window
400k tokensBenchmarks & Scores
Coding (swe-bench-pro)Winner (+8.7%)
54.4%complex codebases, multi-file repositories, and architectural planning
Reasoning (gpqa-diamond)Winner (+5.9%)
87.5%graduate-level science QA
Cost & Context
Cost (per 1M tokens)
$1.69Input: $0.75 | Output: $4.50Context Window
400k tokensFrequently Asked Questions about GPT-5 mini vs GPT-5.4 mini
GPT-5 mini is cheaper than GPT-5.4 mini. GPT-5 mini has a blended cost of $0.69/1M tokens, which is about 2.5x cheaper than GPT-5.4 mini at $1.69/1M tokens.
GPT-5.4 mini is better for coding tasks on this benchmark. It scores 54.4% on swe-bench-pro (complex codebases, multi-file repositories, and architectural planning) compared to GPT-5 mini which scores 45.7%.
Related Matchups
Explore similar comparisons for GPT-5 mini and GPT-5.4 mini.
Do you want to find a model for your constraints?
Use our interactive model finder to filter LLMs by reasoning capability, coding performance, cost, and context length.