Auto model (Standard) vs GPT-5.4 Mini
GPT-5.4 Mini is cheaper on output tokens, while Auto model (Standard) offers a larger context window. Choose Auto model (Standard) or GPT-5.4 Mini based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Auto model (Standard) | GPT-5.4 Mini |
|---|---|---|
| Provider | NanoGPT | OpenAI |
| Input / 1M tokens | $10.00 | $0.75 |
| Output / 1M tokens | $19.99 | $4.50 |
| Context window | 1M | 400K |
| Parameters | — | — |
| Open weights | No | No |
| Released | Apr 2025 | Mar 2026 |
| LiveBench score | — | 67.7 |
| FrontierMath score | — | 28.3 |
| Epoch Capabilities Index score | — | 149 |
| GPQA Diamond score | — | 86.9 |
| LMArena (Chatbot Arena) Elo score | — | 1411 |
| LMArena WebDev Elo score | — | 1398 |
| LMArena Vision Elo score | — | 1245 |
FAQ
Is Auto model (Standard) or GPT-5.4 Mini cheaper?
GPT-5.4 Mini is cheaper on output tokens ($4.50 vs $19.99 per 1M).
Which has the larger context window, Auto model (Standard) or GPT-5.4 Mini?
Auto model (Standard) has the larger context window (1M tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.