GPT-5.6 SolvsGPT-5.4
Across 10 shared benchmarks, GPT-5.6 Sol leads overall: GPT-5.6 Sol wins 9, GPT-5.4 wins 1, with 0 ties and an average score difference of +46.53.
GPT-5.6 Sol
OpenAI · 2026-06-26 · Reasoning model
GPT-5.4
OpenAI · 2026-03-05 · Multimodal model
GPT-5.6 Sol9 wins(90%)(10%)1 winGPT-5.4
Benchmark scores
Grouped by capability, sorted by largest gap within each. 10 shared benchmarks.
Coding and Software Engineer
GPT-5.6 Sol 4/4| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| Text Arena (Coding) | 1,6204 / 35极高强度思考（无工具） | 1,45724 / 35Thinking High (No Tools) | +163.04 |
| WeirdML v2 | 88.762 / 52Thinking High (With Tools) | 57.4431 / 52Normal (With Tools) | +31.32 |
| DeepSWE | 72.701 / 32极高强度思考（工具） | 5224 / 32极高强度思考（工具） | +20.70 |
| SWE-Bench Pro - Public | 64.608 / 59极高强度思考（工具） | 57.7021 / 59极高强度思考（无工具） | +6.90 |
Math and Reasoning
GPT-5.6 Sol 2/2| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| FrontierMath Tier 4 v2 | 82.932 / 40最高（无工具） | 4910 / 40极高强度思考（无工具） | +33.93 |
| FrontierMath v2 | 89.121 / 58最高（无工具） | 78.6010 / 58极高强度思考（无工具） | +10.53 |
General Evaluation
GPT-5.6 Sol 1/1| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| GPQA Diamond | 82.83121 / 270Normal (No Tools) | 74.75173 / 270Normal (No Tools) | +8.08 |
General Knowledge
GPT-5.4 1/1| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| HLE | 49.5038 / 188最高（无工具） | 52.1027 / 188极高强度思考（工具） | -2.60 |
Text Embedding
GPT-5.6 Sol 1/1| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| Context Arena | 97.632 / 126最高（无工具） | 32.79112 / 126Normal (No Tools) | +64.84 |
Writing and Creative Capabilities
GPT-5.6 Sol 1/1| Benchmark | GPT-5.6 Sol | GPT-5.4 | Diff |
|---|---|---|---|
| Creative Writing | 1,9644 / 99Normal (No Tools) | 1,83611 / 99Normal (No Tools) | +128.60 |
Specs
| Field | GPT-5.6 Sol | GPT-5.4 |
|---|---|---|
| Publisher | OpenAI | OpenAI |
| Release date | 2026-06-26 | 2026-03-05 |
| Model type | Reasoning model | Multimodal model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 1.05M | 1M |
| Max output | 128K | 125K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GPT-5.6 Sol | GPT-5.4 |
|---|---|---|
| Text input | $4 / 1M tokens | $2.5 / 1M tokens |
| Text output | $20 / 1M tokens | $15 / 1M tokens |
| Cache read | $0.4 / 1M tokens | $0.25 / 1M tokens |
| Cache write | $5 / 1M tokens | Not public |
Summary
- GPT-5.6 Solleads in:Coding and Software Engineer (4/4), Math and Reasoning (2/2), General Evaluation (1/1), Text Embedding (1/1), Writing and Creative Capabilities (1/1)
- GPT-5.4leads in:General Knowledge (1/1)
On average across the 10 shared benchmarks, GPT-5.6 Sol scores 46.53 higher.
Largest single-benchmark gap: Text Arena (Coding) — GPT-5.6 Sol 1,620 vs GPT-5.4 1,457 (+163.04).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.