Opus 4.7vsGPT-5.4
Across 13 shared benchmarks, GPT-5.4 leads overall: Opus 4.7 wins 5, GPT-5.4 wins 7, with 1 ties and an average score difference of +0.01.
Opus 4.75 wins(38%)Ties1(54%)7 winsGPT-5.4
Benchmark scores
Grouped by capability, sorted by largest gap within each. 13 shared benchmarks.
General Knowledge
GPT-5.4 3/6| Benchmark | Opus 4.7 | GPT-5.4 | Diff |
|---|---|---|---|
| LiveBench | 76.917 / 115Deep Thinking (No Tools) | 80.282 / 115Deep Thinking (No Tools) | -3.37 |
| HLE | 54.7013 / 172Extended (with tools) | 52.1021 / 172极高强度思考（工具） | +2.60 |
| GPQA Diamond | 94.204 / 187Extended (no tools) | 92.8011 / 187极高强度思考（无工具） | +1.40 |
| ARC-AGI-2 | 75.8011 / 62最高（无工具） | 77.109 / 62Normal (No Tools) | -1.30 |
| ARC-AGI | 93.5011 / 68Thinking High (No Tools) | 93.709 / 68Normal (No Tools) | -0.20 |
| ARC-AGI-3 | 08 / 9Thinking High (No Tools) | 07 / 9Thinking High (No Tools) | — |
AI Agent - Tool Usage
Opus 4.7 2/3| Benchmark | Opus 4.7 | GPT-5.4 | Diff |
|---|---|---|---|
| MCP-Atlas | 79.107 / 27Deep Thinking (With Tools) | 70.6014 / 27极高强度思考（工具） | +8.50 |
| Terminal Bench 2.0 | 69.406 / 47Extended (with tools) | 75.104 / 47极高强度思考（工具） | -5.70 |
| OSWorld-Verified | 7810 / 24Extended (with tools) | 7512 / 24极高强度思考（工具） | +3 |
Math and Reasoning
GPT-5.4 2/2| Benchmark | Opus 4.7 | GPT-5.4 | Diff |
|---|---|---|---|
| FrontierMath - Tier 4 | 22.9012 / 80极高强度思考（无工具） | 27.1011 / 80极高强度思考（无工具） | -4.20 |
| FrontierMath | 43.806 / 60极高强度思考（无工具） | 47.605 / 60极高强度思考（无工具） | -3.80 |
AI Agent - Information Search
GPT-5.4 1/1| Benchmark | Opus 4.7 | GPT-5.4 | Diff |
|---|---|---|---|
| BrowseComp | 79.3017 / 53Extended (with tools) | 82.7015 / 53极高强度思考（工具） | -3.40 |
Coding and Software Engineer
Opus 4.7 1/1| Benchmark | Opus 4.7 | GPT-5.4 | Diff |
|---|---|---|---|
| SWE-Bench Pro - Public | 64.307 / 54Extended (with tools) | 57.7017 / 54极高强度思考（无工具） | +6.60 |
Specs
| Field | Opus 4.7 | GPT-5.4 |
|---|---|---|
| Publisher | Anthropic | OpenAI |
| Release date | 2026-04-16 | 2026-03-05 |
| Model type | Reasoning model | Multimodal model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 1000K | 1M |
| Max output | 128K | 125K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Opus 4.7 | GPT-5.4 |
|---|---|---|
| Text input | $5 / 1M tokens | $2.5 / 1M tokens |
| Text output | $25 / 1M tokens | $15 / 1M tokens |
| Cache read | $0.5 / 1M tokens | Not public |
| Cache write | $6.25 / 1M tokens | $0.25 / 1M tokens |
Summary
- Opus 4.7leads in:AI Agent - Tool Usage (2/3), Coding and Software Engineer (1/1)
- GPT-5.4leads in:General Knowledge (3/6), Math and Reasoning (2/2), AI Agent - Information Search (1/1)
On average across the 13 shared benchmarks, Opus 4.7 scores 0.01 higher.
Largest single-benchmark gap: MCP-Atlas — Opus 4.7 79.10 vs GPT-5.4 70.60 (+8.50).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.