# GPT-5.4 Pro vs Step 3.5 Flash

> GPT-5.4 Pro is the stronger model overall, scoring 58.9 to 42.3 on the Noometry Index. Step 3.5 Flash costs 450× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-4-pro-vs-step-3-5-flash
- Last updated: 2026-10-11
- Shared benchmarks: 0

## Summary

- The widest gap is in reasoning, where GPT-5.4 Pro leads 70.7 to 22.2.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $30 / $180 for GPT-5.4 Pro.
- GPT-5.4 Pro accepts more context: 1.05M tokens versus 256K.
- Step 3.5 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| Provider | OpenAI | StepFun |
| Noometry Index | 58.9 | 42.3 |
| Rank | 18 | 116 |
| Context | 1.05M | 256K |
| Input $/M | $30 | $0.10 |
| Output $/M | $180 | $0.30 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.4 Pro: 43.2 (#90)
- Step 3.5 Flash: 42.4 (#105)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| WeirdML | 57.4% | — |
| LMArena Coding | — | 1436 |

## Reasoning

- GPT-5.4 Pro: 70.7 (#13)
- Step 3.5 Flash: 22.2 (#202)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| ARC-AGI-2 | 83.3% | — |
| SimpleBench | 74.1% | — |
| NYT Connections (extended) | — | 28.4% |
| ARC-AGI-1 | 94.5% | — |
| CritPt | 30% | — |
| Chess Puzzles | 58.6% | — |
| EnigmaEval | 23.8% | — |
| LMArena Hard Prompts | — | 1411 |
| Epoch Capabilities Index | 158.93 | — |

## Math

- GPT-5.4 Pro: 72.4 (#22)
- Step 3.5 Flash: 42.6 (#84)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.5% | — |
| FrontierMath Tier 4 | 58.5% | — |
| MathArena Final-Answer Competitions | — | 66.8% |
| LMArena Math | — | 1408 |
| FrontierMath (Feb 2025 set) | 50% | — |
| FrontierMath Tier 4 (v1) | 37.5% | — |

## Knowledge

- GPT-5.4 Pro: 68.3 (#7)
- Step 3.5 Flash: 39.6 (#132)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| GPQA Diamond | 94.6% | — |
| Humanity's Last Exam | 44.3% | — |
| SimpleQA Verified | 46.3% | — |
| Vectara Hallucination Rate | 8.3% | — |
| LMArena Expert | — | 1421 |

## Multilingual

- GPT-5.4 Pro: —
- Step 3.5 Flash: 50.5 (#119)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | — | 1385 |
| LMArena Chinese | — | 1447 |
| LMArena French | — | 1421 |
| LMArena German | — | 1405 |
| LMArena Japanese | — | 1354 |
| LMArena Korean | — | 1352 |
| LMArena Russian | — | 1385 |
| LMArena Spanish | — | 1419 |

## Instruction Following

- GPT-5.4 Pro: —
- Step 3.5 Flash: 73.1 (#124)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | — | 1385 |

## Long Context

- GPT-5.4 Pro: —
- Step 3.5 Flash: 42.8 (#117)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | — | 1402 |

## Writing & Preference

- GPT-5.4 Pro: —
- Step 3.5 Flash: 58.8 (#113)

| Benchmark | GPT-5.4 Pro | Step 3.5 Flash |
|---|---|---|
| LMArena Text | — | 1403 |
| LMArena Creative Writing | — | 1357 |
| LMArena Multi-Turn | — | 1405 |

## FAQ

### Is GPT-5.4 Pro better than Step 3.5 Flash?

GPT-5.4 Pro is the stronger model overall, scoring 58.9 to 42.3 on the Noometry Index. Step 3.5 Flash costs 450× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.4 Pro or Step 3.5 Flash?

Step 3.5 Flash is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; GPT-5.4 Pro lists at $30 and $180.

### Is GPT-5.4 Pro or Step 3.5 Flash better for coding?

They score almost the same on coding (43.2 vs 42.4); test both on your own repository before choosing.

### Which has the bigger context window?

GPT-5.4 Pro does, with 1.05M tokens against 256K.

### How many benchmarks do GPT-5.4 Pro and Step 3.5 Flash share?

0 benchmarks have published results for both models. GPT-5.4 Pro has 16 scored results on Noometry and Step 3.5 Flash has 19.