Model comparison
Claude Opus 4.8 vs Claude Sonnet 5
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.6 on the Noometry Index. Claude Sonnet 5 costs 2.5× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.
Last verified . 50 shared benchmarks.
Summary
- They share 50 benchmarks with published results for both. Claude Opus 4.8 scores higher in 10 categories and Claude Sonnet 5 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.8 leads 64.7 to 49.1.
- The biggest single-benchmark swing is Surface Evolver Bench: 87.5% for Claude Opus 4.8 and 60% for Claude Sonnet 5.
- Claude Sonnet 5 is cheaper at $2 / $10 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.
Side by side
| Claude Opus 4.8 | Claude Sonnet 5 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 60.7 | 54.6 |
| Released | 2026-05-28 | 2026-06-29 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1M |
| Max output | 128K | 128K |
| Input $ / M tokens | $5 | $2 |
| Output $ / M tokens | $25 | $10 |
| Results tracked | 65 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.8 leads
Claude Opus 4.8: 59.9 (#12), Claude Sonnet 5: 55.5 (#26)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| DeepSWE | 59% | 53.8% |
| FrontierCode | 46.5% | 42.7% |
| LMArena WebDev | 1556 | 1541 |
| SciCode | 53.5% | 54.3% |
| GSO | 47.1% | 37.3% |
| WeirdML | 82.9% | 68.8% |
| LMArena Coding | 1490 | 1483 |
| ALE-Bench | 1,564 | 1,463 |
| CursorBench | — | 34.1% |
Agentic & Tool Use Claude Opus 4.8 leads
Claude Opus 4.8: 47.6 (#11), Claude Sonnet 5: 42.8 (#18)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| APEX-Agents | 48.9% | 54.5% |
| GBAEval | 70.9% | 65.3% |
| LMArena Search | 1204 | 1194 |
| Vending-Bench 2 | 5,787 | 6,378 |
| OSWorld 2.0 | 20.6% | — |
| Remote Labor Index | 8.3% | — |
| τ²-bench Banking | 39.7% | — |
| DeepResearch Bench | 50.2% | — |
| PostTrainBench | 33.8% | — |
| GDP.pdf | 24% | — |
Reasoning Claude Opus 4.8 leads
Claude Opus 4.8: 64.7 (#16), Claude Sonnet 5: 49.1 (#39)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| SimpleBench | 64.8% | 60.6% |
| NYT Connections (extended) | 91.1% | 75.1% |
| CritPt | 20.9% | 16.9% |
| Chess Puzzles | 34% | 35% |
| LMArena Hard Prompts | 1482 | 1461 |
| Mystery Game Puzzles | 36% | 35% |
| DTBench | 94.9% | 92.5% |
| LMCA | 57.5% | 50% |
| Surface Evolver Bench | 87.5% | 60% |
| Bench to the Future 3 | 0.14 | 0.14 |
| Epoch Capabilities Index | 158.21 | 156.21 |
| ForecastBench | 59.9 | 61.1 |
| ARC-AGI-2 | 72.1% | — |
| Kagi LLM Benchmark | 88.8% | — |
| ARC-AGI-1 | 92.5% | — |
| EnigmaEval | 23.5% | — |
| EBR-Bench | 28.6% | — |
Math Claude Opus 4.8 leads
Claude Opus 4.8: 78.4 (#13), Claude Sonnet 5: 66.2 (#27)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 80% | 65.6% |
| FrontierMath Tier 4 | 56.1% | 29.3% |
| OTIS Mock AIME 2024-2025 | 98.3% | 94.7% |
| ProofBench | 69% | 77% |
| LMArena Math | 1487 | 1467 |
| MathArena Final-Answer Competitions | 91.8% | — |
| FrontierMath (Feb 2025 set) | 47.2% | — |
| FrontierMath Tier 4 (v1) | 31.3% | — |
Knowledge Claude Opus 4.8 leads
Claude Opus 4.8: 61.3 (#29), Claude Sonnet 5: 55.6 (#47)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| GPQA Diamond | 91% | 90.5% |
| SimpleQA Verified | 53% | 33.7% |
| LMArena Expert | 1502 | 1490 |
Multimodal Too close to call
Claude Opus 4.8: 42.9 (#26), Claude Sonnet 5: 42.4 (#31)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| LMArena Vision | 1294 | 1274 |
| Blueprint-Bench 2 | 14.5% | 24.9% |
| LMArena Document | 1475 | 1466 |
| Furniture Assembly | 42.5% | — |
Multilingual Claude Opus 4.8 leads
Claude Opus 4.8: 55.2 (#33), Claude Sonnet 5: 53.8 (#55)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| LMArena Non-English | 1450 | 1431 |
| LMArena Chinese | 1507 | 1477 |
| LMArena French | 1481 | 1460 |
| LMArena German | 1472 | 1440 |
| LMArena Japanese | 1440 | 1422 |
| LMArena Korean | 1432 | 1411 |
| LMArena Russian | 1474 | 1451 |
| LMArena Spanish | 1466 | 1437 |
Instruction Following Claude Opus 4.8 leads
Claude Opus 4.8: 77.4 (#24), Claude Sonnet 5: 76.3 (#41)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| LMArena Instruction Following | 1476 | 1452 |
Long Context Too close to call
Claude Opus 4.8: 45.4 (#35), Claude Sonnet 5: 44.8 (#55)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| LMArena Longer Query | 1483 | 1463 |
Writing & Preference Claude Opus 4.8 leads
Claude Opus 4.8: 72.0 (#16), Claude Sonnet 5: 69.2 (#25)
| Benchmark | Claude Opus 4.8 | Claude Sonnet 5 |
|---|---|---|
| LMArena Text | 1461 | 1442 |
| LMArena Creative Writing | 1454 | 1416 |
| EQ-Bench Creative Writing | 1840 | 1794 |
| EQ-Bench 4 | 1281 | 1236 |
| LMArena Multi-Turn | 1476 | 1454 |
Frequently asked questions
Is Claude Opus 4.8 better than Claude Sonnet 5?
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.6 on the Noometry Index. Claude Sonnet 5 costs 2.5× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 4.8 or Claude Sonnet 5?
Claude Sonnet 5 is cheaper. It lists at $2 per million input tokens and $10 per million output tokens; Claude Opus 4.8 lists at $5 and $25.
Is Claude Opus 4.8 or Claude Sonnet 5 better for coding?
Claude Opus 4.8 scores higher on coding benchmarks: 59.9 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do Claude Opus 4.8 and Claude Sonnet 5 share?
50 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and Claude Sonnet 5 has 51.