Compare models. Find the differences.
Add, remove, or change models in the table.
Benchmark snapshot
0 benchmarks in common.
4 models selected. 0 benchmarks in common.
Reported scores; test settings may differ.
| Benchmark | DeepSeek | DeepSeek | Qwen | Anthropic |
|---|---|---|---|---|
| Arena ELOHuman preference | 1,476 | 1,465 | Not in our dataset | 1,503 |
| LiveBench 1.0General reasoning | 81.1% | Not in our dataset | Not in our dataset | 74.5% |
| GPQA DiamondScientific knowledge | Not in our dataset | 90.1% | Not in our dataset | 91.3% |
| HLEAcademic reasoning | Not in our dataset | 37.7% | Not in our dataset | 40.0% |
| MMLU ProMulti-task understanding | Not in our dataset | 87.5% | Not in our dataset | Not in our dataset |
| SimpleBenchCommonsense reasoning | 66.7% | Not in our dataset | 62.5% | 67.6% |
| AIME 2025Mathematics | Not in our dataset | Not in our dataset | Not in our dataset | 95.6% |
| GSOGraduate science | Not in our dataset | Not in our dataset | Not in our dataset | 41.2% |
| SWE-Bench VerifiedAgentic coding | Not in our dataset | 80.6% | Not in our dataset | 80.8% |
| LiveCodeBenchCompetitive coding | Not in our dataset | 89.8% | Not in our dataset | Not in our dataset |
| Tau-BenchAgentic tool use | Not in our dataset | 88.9% | Not in our dataset | 91.9% |
| Terminal-BenchAgentic terminal | Not in our dataset | Not in our dataset | Not in our dataset | 65.4% |
| ARC-AGI-2Visual reasoning | 67.5% | 61.3% | Not in our dataset | 68.8% |
Pricing & context Published USD rates per 1M tokens
| Specification | V4.1 Flash Thinking | V4 Pro Thinking | 3.8 2.4T A95B Thinking | Claude Opus 4 4.6 Thinking |
|---|---|---|---|---|
| Input | $0.045 | $1.6 | $2 | $5 |
| Output | $0.5 | $3.2 | $6 | $25 |
| Context window | 1.0M | 1.0M | 1.0M | 1.0M |
| 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.