Model comparison
Claude Opus 4 vs o4-mini
Claude Opus 4 is the stronger model overall, scoring 42.9 to 41.6 on the Noometry Index. o4-mini costs 16× less per token, which makes it the better buy when Claude Opus 4's lead doesn't matter for your workload.
Last verified . 51 shared benchmarks.
Summary
- They share 51 benchmarks with published results for both. Claude Opus 4 scores higher in 7 categories and o4-mini in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where o4-mini leads 39.5 to 30.9.
- The biggest single-benchmark swing is ARC-AGI-1: 35.7% for Claude Opus 4 and 58.7% for o4-mini.
- o4-mini is cheaper at $1.10 / $4.40 per million input/output tokens, against $15 / $75 for Claude Opus 4.
Side by side
| Claude Opus 4 | o4-mini | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 42.9 | 41.6 |
| Released | 2025-05-22 | 2025-04-16 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 200K |
| Max output | 32K | 100K |
| Input $ / M tokens | $15 | $1.10 |
| Output $ / M tokens | $75 | $4.40 |
| Results tracked | 56 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4 leads
Claude Opus 4: 47.8 (#55), o4-mini: 41.6 (#110)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| SWE-bench Verified (bash only) | 67.6% | 45% |
| Aider Polyglot | 72% | 72% |
| GSO | 6.9% | 3.6% |
| WeirdML | 43.7% | 52.6% |
| LMArena Coding | 1442 | 1369 |
| AlgoTune | 1.33 | 1.72 |
| SWE-bench Verified | 70.7% | — |
| CadEval | — | 62% |
| ALE-Bench | — | 826.17 |
Agentic & Tool Use Claude Opus 4 leads
Claude Opus 4: 34.5 (#39), o4-mini: 32.3 (#57)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| METR Time Horizons | 63.9% | 63.9% |
| Berkeley Function Calling Leaderboard | — | 53.2% |
| GDPval | — | 25.3% |
| Cybench | 38% | — |
| DeepResearch Bench | 46.8% | — |
| LMArena Search | 1127 | — |
Reasoning Claude Opus 4 leads
Claude Opus 4: 27.4 (#104), o4-mini: 24.6 (#143)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| ARC-AGI-2 | 8.6% | 6.1% |
| SimpleBench | 58.8% | 38.7% |
| Kagi LLM Benchmark | 74.3% | 67.6% |
| ARC-AGI-1 | 35.7% | 58.7% |
| CritPt | 0.3% | 0.6% |
| EnigmaEval | 5.6% | 9.2% |
| LMArena Hard Prompts | 1399 | 1351 |
| DTBench | 81.6% | 77.6% |
| LMCA | 37.4% | 26.5% |
| Epoch Capabilities Index | 142.67 | 145.63 |
| ForecastBench | 61.1 | 61.8 |
| Chess Puzzles | — | 26% |
| Mystery Game Puzzles | — | 5% |
Math Too close to call
Claude Opus 4: 40.2 (#80), o4-mini: 40.2 (#79)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 81.7% |
| Omni-MATH | 61.6% | 72% |
| LMArena Math | 1391 | 1390 |
| MATH Level 5 | 85% | 97.8% |
| FrontierMath (Feb 2025 set) | 4.5% | 24.8% |
| FrontierMath Tier 4 (v1) | 4.2% | 6.3% |
| FrontierMath (Tiers 1-3) | — | 36.1% |
| FrontierMath Tier 4 | — | 4.9% |
Knowledge Too close to call
Claude Opus 4: 43.9 (#82), o4-mini: 43.5 (#83)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| GPQA Diamond | 76.3% | 79.6% |
| Humanity's Last Exam | 10.7% | 18.1% |
| MMLU-Pro | 87.5% | 82% |
| Confabulations | 15.9% | 15.8% |
| Vectara Hallucination Rate | 12% | 18.6% |
| GPQA (HELM) | 70.8% | 73.5% |
| LMArena Expert | 1384 | 1344 |
| SimpleQA Verified | — | 19.6% |
Multimodal o4-mini leads
Claude Opus 4: 30.9 (#101), o4-mini: 39.5 (#46)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| LMArena Vision | 1192 | 1194 |
| GeoBench | 49% | 64% |
| VPCT | 38% | 57.5% |
Multilingual Claude Opus 4 leads
Claude Opus 4: 49.0 (#133), o4-mini: 47.2 (#149)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| LMArena Non-English | 1362 | 1336 |
| LMArena Chinese | 1386 | 1355 |
| LMArena French | 1370 | 1362 |
| LMArena German | 1391 | 1336 |
| LMArena Japanese | 1332 | 1308 |
| LMArena Korean | 1318 | 1313 |
| LMArena Russian | 1392 | 1334 |
| LMArena Spanish | 1388 | 1346 |
Instruction Following Claude Opus 4 leads
Claude Opus 4: 77.3 (#26), o4-mini: 75.4 (#63)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| IFEval | 91.8% | 92.8% |
| LMArena Instruction Following | 1406 | 1321 |
Long Context o4-mini leads
Claude Opus 4: 39.5 (#167), o4-mini: 45.5 (#31)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| Fiction.LiveBench | 61.1% | 77.8% |
| LMArena Longer Query | 1422 | 1315 |
Writing & Preference Claude Opus 4 leads
Claude Opus 4: 61.4 (#85), o4-mini: 54.2 (#148)
| Benchmark | Claude Opus 4 | o4-mini |
|---|---|---|
| LMArena Text | 1377 | 1353 |
| LMArena Creative Writing | 1386 | 1294 |
| Short-Story Creative Writing | 83.6% | 75% |
| WildBench | 85.2% | 85.4% |
| LMArena Multi-Turn | 1396 | 1350 |
| EQ-Bench Creative Writing | 1580 | — |
Frequently asked questions
Is Claude Opus 4 better than o4-mini?
Claude Opus 4 is the stronger model overall, scoring 42.9 to 41.6 on the Noometry Index. o4-mini costs 16× less per token, which makes it the better buy when Claude Opus 4's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 4 or o4-mini?
o4-mini is cheaper. It lists at $1.10 per million input tokens and $4.40 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or o4-mini better for coding?
Claude Opus 4 scores higher on coding benchmarks: 47.8 versus 41.6 in the Noometry coding category.
Which has the bigger context window?
Both accept 200K tokens.
How many benchmarks do Claude Opus 4 and o4-mini share?
51 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and o4-mini has 60.