GPT-5.4 nano vs GPT-5 nano
| Spec | GPT-5.4 nano | GPT-5 nano |
|---|---|---|
| Respan Index | 24.1 (#109) | 18.8 (#138) |
| Blended price per 1M | $0.463 | $0.138 |
| Context window | 400K | 400K |
| Input / output per 1M | $0.2 / $1.25 | $0.05 / $0.4 |
| Max output | 128K | 128K |
| Input | text, vision | text, vision |
| Released | 2026-03-17 | 2025-08-07 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | GPT-5.4 nano | GPT-5 nano |
|---|---|---|
| Coding | 9.0 GPT-5.4 nano | 4.2 GPT-5 nano |
| Agents | 22.3 GPT-5.4 nano | 18.5* GPT-5 nano |
| Reasoning | 32.4 GPT-5.4 nano | 25.3 GPT-5 nano |
| Knowledge & language | 13.2 GPT-5.4 nano | 11.7 GPT-5 nano |
| Human preference | 45.6 GPT-5.4 nano | 37.6 GPT-5 nano |
* Estimated: no published scores in that category. How the Respan Index works
Benchmarks both report
GPT-5.4 nano leads on 9 of the 11 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | GPT-5.4 nano | GPT-5 nano |
|---|---|---|
| LMArena Elo | 1401 | 1338 |
| LMArena Vision (LMArena) | 1199 | 1148 |
| Chess Puzzles (Epoch AI) | 45% | 22% |
| FrontierMath Tier 4 v2 (Epoch AI) | 12.2% | 2.4% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 44.9% | 20% |
| GPQA Diamond | 82.8% | 69.4% |
| Mystery Game Puzzles (Epoch AI) | 9% | 9% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 87.8% | 81.1% |
| SAGE (Vals AI) | 38.1% | 30.4% |
| SimpleQA Verified | 11.7% | 11.7% |
| WeirdML (Håvard Tveit Ihle) | 49.2% | 38.1% |
FAQ
Is GPT-5.4 nano better than GPT-5 nano?
GPT-5.4 nano scores 24.1 on the Respan Index (#109), ahead of GPT-5 nano at 18.8 (#138). GPT-5.4 nano leads on 9 of the 11 benchmarks both report.
Which is cheaper, GPT-5.4 nano or GPT-5 nano?
GPT-5.4 nano lists at $0.2 input and $1.25 output per 1M tokens; GPT-5 nano at $0.05 and $0.4. GPT-5 nano is 3.4x cheaper per token at list price.
Which has the larger context window?
Both accept 400K tokens of context.