Compare models. Find the differences.
Add, remove, or change models in the table.
Benchmark snapshot
5 benchmarks in common.
4 models selected. 5 benchmarks in common.
Reported scores; test settings may differ.
| Benchmark | OpenAI | DeepSeek | DeepSeek | OpenAI |
|---|---|---|---|---|
| Arena ELOHuman preference | 1,391 | 1,421 | 1,458 | 1,390 |
| LiveBench 1.0General reasoning | Not in our dataset | Not in our dataset | 77.4% | Not in our dataset |
| GPQA DiamondScientific knowledge | 81.4% | 81.0% | 72.9% | 82.3% |
| HLEAcademic reasoning | 18.1% | Not in our dataset | 7.7% | 16.7% |
| MMLU ProMulti-task understanding | Not in our dataset | 85.0% | 82.9% | 83.7% |
| SimpleBenchCommonsense reasoning | 38.7% | 40.8% | 50.9% | Not in our dataset |
| AIME 2025Mathematics | 92.7% | 87.5% | Not in our dataset | 91.1% |
| GSOGraduate science | 3.6% | Not in our dataset | Not in our dataset | Not in our dataset |
| AiderCode editing | 72.0% | 71.6% | Not in our dataset | Not in our dataset |
| SWE-Bench VerifiedAgentic coding | 68.1% | 57.6% | 73.6% | 72.0% |
| LiveCodeBenchCompetitive coding | 80.2% | 73.1% | 56.8% | 80.4% |
| Tau-BenchAgentic tool use | 65.6% | 63.9% | Not in our dataset | Not in our dataset |
| Terminal-BenchAgentic terminal | Not in our dataset | Not in our dataset | Not in our dataset | 24.0% |
| ARC-AGI-2Visual reasoning | 6.1% | 1.1% | 0.8% | 4.4% |
Pricing & context Published USD rates per 1M tokens
| Specification | o4 Mini o4-mini | R1 DeepSeek-R1-0528 | V4 Pro | GPT-5 Mini Thinking (5.0) |
|---|---|---|---|---|
| Input | $1.1 | $0.5 | $1.6 | $0.25 |
| Output | $4.4 | $2.15 | $3.2 | $2 |
| Context window | 200K | 164K | 1.0M | 400K |
| Pricing details | Explore pricing → | Explore pricing → | Explore pricing → | Explore pricing → |
How to read this comparison
Each column keeps the exact model version you selected. Shared benchmarks are the intersection of their reported results; “all coverage” also shows where evidence is missing. Missing results never count as zero.
These are published benchmark scores, not a controlled head-to-head test. The available evidence does not establish matching evaluation settings, so we show no cross-model winner highlights or deltas. Composite quality scores are omitted because they can summarize different benchmark sets.
Scores and statistics are computed upstream. This page reads a fixed data snapshot; its date describes the snapshot, not the date of every measurement.