GPT-5vsGPT-4.5
Across 4 shared benchmarks, GPT-5 leads overall: GPT-5 wins 4, GPT-4.5 wins 0, with 0 ties and an average score difference of +29.
GPT-54 wins(100%)(0%)0 winsGPT-4.5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
Agent Level Benchmark
GPT-5 1/1| Benchmark | GPT-5 | GPT-4.5 | Diff |
|---|---|---|---|
| Aider-Polyglot | 881 / 59Thinking High (No Tools) | 44.9037 / 59Normal (No Tools) | +43.10 |
Coding and Software Engineer
GPT-5 1/1| Benchmark | GPT-5 | GPT-4.5 | Diff |
|---|---|---|---|
| SWE-bench Verified | 72.8050 / 112 | 38104 / 112 | +34.80 |
Commonsense Reasoning
GPT-5 1/1| Benchmark | GPT-5 | GPT-4.5 | Diff |
|---|---|---|---|
| Simple Bench | 56.7020 / 63Thinking High (No Tools) | 34.5046 / 63Normal (No Tools) | +22.20 |
General Knowledge
GPT-5 1/1| Benchmark | GPT-5 | GPT-4.5 | Diff |
|---|---|---|---|
| GPQA Diamond | 87.3040 / 187 | 71.40111 / 187 | +15.90 |
Specs
| Field | GPT-5 | GPT-4.5 |
|---|---|---|
| Publisher | OpenAI | OpenAI |
| Release date | 2025-08-07 | 2025-02-28 |
| Model type | Foundation model | Chat model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 400K | 128K |
| Max output | 128K | Not available |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GPT-5 | GPT-4.5 |
|---|---|---|
| Text input | $1.25 / 1M tokens | Not public |
| Text output | $10 / 1M tokens | Not public |
| Cache read | $0.125 / 1M tokens | Not public |
| Cache write | $0 / 1M tokens | Not public |
One or both models have incomplete public pricing.
Summary
- GPT-5leads in:Agent Level Benchmark (1/1), Coding and Software Engineer (1/1), Commonsense Reasoning (1/1), General Knowledge (1/1)
On average across the 4 shared benchmarks, GPT-5 scores 29 higher.
Largest single-benchmark gap: Aider-Polyglot — GPT-5 88 vs GPT-4.5 44.90 (+43.10).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.