GPT-5.2vsOpus 4.5
Across 10 shared benchmarks, Opus 4.5 leads overall: GPT-5.2 wins 3, Opus 4.5 wins 7, with 0 ties and an average score difference of -8.44.
GPT-5.23 wins(30%)(70%)7 winsOpus 4.5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 10 shared benchmarks.
Agent Level Benchmark
Opus 4.5 2/2| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| τ²-Bench - Telecom | 46.50178 / 264Normal (With Tools) | 86.3082 / 264Normal (With Tools) | -39.80 |
| Terminal Bench Hard | 31.80100 / 244Normal (With Tools) | 40.9050 / 244Normal (With Tools) | -9.10 |
Coding and Software Engineer
Opus 4.5 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| LiveCodeBench | 66.90103 / 250Normal (No Tools) | 73.8074 / 250Normal (No Tools) | -6.90 |
General Evaluation
Opus 4.5 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| GPQA Diamond | 73.23298 / 462Normal (No Tools) | 81220 / 462Normal (No Tools) | -7.77 |
General Knowledge
GPT-5.2 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| CritPt | 0.60168 / 200Normal (No Tools) | 0.30182 / 200Normal (No Tools) | +0.30 |
Instruction Following
GPT-5.2 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| IF Bench | 47.40168 / 282Normal (No Tools) | 43194 / 282Normal (No Tools) | +4.40 |
Long Context
Opus 4.5 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| AA-LCR | 45.70144 / 170Normal (No Tools) | 70.7099 / 170Normal (No Tools) | -25 |
Math and Reasoning
Opus 4.5 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| AIME2025 | 51153 / 215Normal (No Tools) | 62.70133 / 215Normal (No Tools) | -11.70 |
Multimodal Understanding
Opus 4.5 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| MMMU-Pro | 65.80148 / 227Normal (No Tools) | 71.20120 / 227Normal (No Tools) | -5.40 |
Writing and Creative Capabilities
GPT-5.2 1/1| Benchmark | GPT-5.2 | Opus 4.5 | Diff |
|---|---|---|---|
| Creative Writing | 1,70026 / 106Normal (No Tools) | 1,68329 / 106Normal (No Tools) | +16.60 |
Specs
| Field | GPT-5.2 | Opus 4.5 |
|---|---|---|
| Publisher | OpenAI | Anthropic |
| Release date | 2025-12-11 | 2025-11-25 |
| Model type | Chat model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 400K | 200K |
| Max output | Not available | 64K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GPT-5.2 | Opus 4.5 |
|---|---|---|
| Text input | $1.75 / 1M tokens | $5 / 1M tokens |
| Text output | $14 / 1M tokens | $25 / 1M tokens |
| Cache read | $0.175 / 1M tokens | $0.5 / 1M tokens |
| Cache write | $1.75 / 1M tokens | $6.25 / 1M tokens |
Summary
- GPT-5.2leads in:General Knowledge (1/1), Instruction Following (1/1), Writing and Creative Capabilities (1/1)
- Opus 4.5leads in:Agent Level Benchmark (2/2), Coding and Software Engineer (1/1), General Evaluation (1/1), Long Context (1/1), Math and Reasoning (1/1), Multimodal Understanding (1/1)
On average across the 10 shared benchmarks, Opus 4.5 scores 8.44 higher.
Largest single-benchmark gap: τ²-Bench - Telecom — GPT-5.2 46.50 vs Opus 4.5 86.30 (-39.80).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.