Model comparison
GPT-5.3 Codex vs GPT-5.4 nano
GPT-5.3 Codex is the stronger model overall, scoring 45.8 to 41.9 on the Noometry Index. GPT-5.4 nano costs 10× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.
Last verified . 3 shared benchmarks.
Summary
- They share 3 benchmarks with published results for both. GPT-5.3 Codex scores higher in 1 category and GPT-5.4 nano in 0 categories; one gap is clear of the uncertainty.
- The widest gap is in coding, where GPT-5.3 Codex leads 48.6 to 43.6.
- The biggest single-benchmark swing is WeirdML: 79.3% for GPT-5.3 Codex and 49.2% for GPT-5.4 nano.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $1.75 / $14 for GPT-5.3 Codex.
Side by side
| GPT-5.3 Codex | GPT-5.4 nano | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.8 | 41.9 |
| Released | 2026-02-05 | 2026-03-17 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 400K |
| Max output | 128K | 128K |
| Input $ / M tokens | $1.75 | $0.20 |
| Output $ / M tokens | $14 | $1.25 |
| Results tracked | 8 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.3 Codex leads
GPT-5.3 Codex: 48.6 (#56), GPT-5.4 nano: 43.6 (#84)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| WeirdML | 79.3% | 49.2% |
| ALE-Bench | 1,655 | 1,005 |
| SWE-bench Verified | 74.8% | — |
| LMArena WebDev | 1409 | — |
| SciCode | — | 46.9% |
| LMArena Coding | — | 1405 |
Agentic & Tool Use Not comparable
GPT-5.3 Codex: 48.0 (#9), GPT-5.4 nano: —
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| Terminal-Bench | 78.4% | — |
| METR Time Horizons | 74.5% | — |
| Vending-Bench 2 | 5,940 | — |
Reasoning Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 23.7 (#173)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| Epoch Capabilities Index | 156.77 | 145.81 |
| ARC-AGI-2 | — | 5.7% |
| Kagi LLM Benchmark | — | 39.7% |
| ARC-AGI-1 | — | 51.5% |
| CritPt | — | 9.3% |
| Chess Puzzles | — | 30% |
| LMArena Hard Prompts | — | 1381 |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LMCA | — | 36.9% |
| ForecastBench | — | 57.3 |
Math Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 40.9 (#88)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 44.9% |
| FrontierMath Tier 4 | — | 12.2% |
| OTIS Mock AIME 2024-2025 | — | 87.8% |
| ProofBench | — | 5% |
| LMArena Math | — | 1406 |
| FrontierMath (Feb 2025 set) | — | 25.9% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 41.9 (#103)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| GPQA Diamond | — | 78.5% |
| SimpleQA Verified | — | 11.7% |
| Vectara Hallucination Rate | — | 3.1% |
| LMArena Expert | — | 1396 |
Multimodal Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 36.7 (#78)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| LMArena Vision | — | 1196 |
Multilingual Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 48.6 (#140)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| LMArena Non-English | — | 1359 |
| LMArena Chinese | — | 1392 |
| LMArena French | — | 1396 |
| LMArena German | — | 1367 |
| LMArena Japanese | — | 1343 |
| LMArena Korean | — | 1320 |
| LMArena Russian | — | 1363 |
| LMArena Spanish | — | 1371 |
Instruction Following Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 71.9 (#144)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| LMArena Instruction Following | — | 1362 |
Long Context Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 41.6 (#137)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| LMArena Longer Query | — | 1366 |
Writing & Preference Not comparable
GPT-5.3 Codex: —, GPT-5.4 nano: 55.7 (#142)
| Benchmark | GPT-5.3 Codex | GPT-5.4 nano |
|---|---|---|
| LMArena Text | — | 1372 |
| LMArena Creative Writing | — | 1314 |
| LMArena Multi-Turn | — | 1382 |
Frequently asked questions
Is GPT-5.3 Codex better than GPT-5.4 nano?
GPT-5.3 Codex is the stronger model overall, scoring 45.8 to 41.9 on the Noometry Index. GPT-5.4 nano costs 10× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.
Which is cheaper, GPT-5.3 Codex or GPT-5.4 nano?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; GPT-5.3 Codex lists at $1.75 and $14.
Is GPT-5.3 Codex or GPT-5.4 nano better for coding?
GPT-5.3 Codex scores higher on coding benchmarks: 48.6 versus 43.6 in the Noometry coding category.
Which has the bigger context window?
Both accept 400K tokens.
How many benchmarks do GPT-5.3 Codex and GPT-5.4 nano share?
3 benchmarks have published results for both models. GPT-5.3 Codex has 8 scored results on Noometry and GPT-5.4 nano has 40.