Model comparison
Claude Opus 4.8 vs DeepSeek V4 Pro
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.3 on the Noometry Index. DeepSeek V4 Pro costs 10× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.
Last verified . 45 shared benchmarks.
DeepSeekSummary
- They share 45 benchmarks with published results for both. Claude Opus 4.8 scores higher in 9 categories and DeepSeek V4 Pro in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Opus 4.8 leads 47.6 to 32.8.
- The biggest single-benchmark swing is Surface Evolver Bench: 87.5% for Claude Opus 4.8 and 40% for DeepSeek V4 Pro.
- DeepSeek V4 Pro is cheaper at $0.66 / $1.98 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| Claude Opus 4.8 | DeepSeek V4 Pro | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 60.7 | 54.3 |
| Released | 2026-05-28 | 2026-04-24 |
| Weights | Proprietary | Open |
| Context window | 1M | 1M |
| Max output | 128K | 393K |
| Input $ / M tokens | $5 | $0.66 |
| Output $ / M tokens | $25 | $1.98 |
| Results tracked | 65 | 48 |
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), DeepSeek V4 Pro: 52.4 (#34)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| FrontierCode | 46.5% | 28.6% |
| LMArena WebDev | 1556 | 1582 |
| SciCode | 53.5% | 51% |
| WeirdML | 82.9% | 66.2% |
| LMArena Coding | 1490 | 1470 |
| ALE-Bench | 1,564 | 1,403 |
| SWE-bench Verified | — | 77.6% |
| DeepSWE | 59% | — |
| GSO | 47.1% | — |
Agentic & Tool Use Claude Opus 4.8 leads
Claude Opus 4.8: 47.6 (#11), DeepSeek V4 Pro: 32.8 (#58)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| APEX-Agents | 48.9% | 47.3% |
| Vending-Bench 2 | 5,787 | 3,285 |
| OSWorld 2.0 | 20.6% | — |
| Remote Labor Index | 8.3% | — |
| τ²-bench Banking | 39.7% | — |
| DeepResearch Bench | 50.2% | — |
| PostTrainBench | 33.8% | — |
| GBAEval | 70.9% | — |
| GDP.pdf | 24% | — |
| LMArena Search | 1204 | — |
Reasoning Claude Opus 4.8 leads
Claude Opus 4.8: 64.7 (#16), DeepSeek V4 Pro: 56.5 (#24)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| ARC-AGI-2 | 72.1% | 61.3% |
| Kagi LLM Benchmark | 88.8% | 53.5% |
| NYT Connections (extended) | 91.1% | 91.3% |
| ARC-AGI-1 | 92.5% | 90.5% |
| CritPt | 20.9% | 18% |
| Chess Puzzles | 34% | 47% |
| LMArena Hard Prompts | 1482 | 1461 |
| Mystery Game Puzzles | 36% | 43% |
| DTBench | 94.9% | 93.9% |
| LMCA | 57.5% | 45.5% |
| Surface Evolver Bench | 87.5% | 40% |
| Epoch Capabilities Index | 158.21 | 155.31 |
| ForecastBench | 59.9 | 56.1 |
| SimpleBench | 64.8% | — |
| EnigmaEval | 23.5% | — |
| EBR-Bench | 28.6% | — |
| Bench to the Future 3 | 0.14 | — |
Math Claude Opus 4.8 leads
Claude Opus 4.8: 78.4 (#13), DeepSeek V4 Pro: 64.8 (#30)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| FrontierMath (Tiers 1-3) | 80% | 64.6% |
| FrontierMath Tier 4 | 56.1% | 26.8% |
| MathArena Final-Answer Competitions | 91.8% | 76.6% |
| OTIS Mock AIME 2024-2025 | 98.3% | 98.6% |
| ProofBench | 69% | 50% |
| LMArena Math | 1487 | 1455 |
| 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), DeepSeek V4 Pro: 59.5 (#31)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| GPQA Diamond | 91% | 91.7% |
| SimpleQA Verified | 53% | 52.9% |
| LMArena Expert | 1502 | 1464 |
| Vectara Hallucination Rate | — | 8.6% |
Multimodal Not comparable
Claude Opus 4.8: 42.9 (#26), DeepSeek V4 Pro: —
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Vision | 1294 | — |
| Blueprint-Bench 2 | 14.5% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1475 | — |
Multilingual Too close to call
Claude Opus 4.8: 55.2 (#33), DeepSeek V4 Pro: 54.4 (#45)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Non-English | 1450 | 1439 |
| LMArena Chinese | 1507 | 1486 |
| LMArena French | 1481 | 1472 |
| LMArena German | 1472 | 1458 |
| LMArena Japanese | 1440 | 1445 |
| LMArena Korean | 1432 | 1447 |
| LMArena Russian | 1474 | 1453 |
| LMArena Spanish | 1466 | 1458 |
Instruction Following Claude Opus 4.8 leads
Claude Opus 4.8: 77.4 (#24), DeepSeek V4 Pro: 76.1 (#47)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Instruction Following | 1476 | 1448 |
Long Context Too close to call
Claude Opus 4.8: 45.4 (#35), DeepSeek V4 Pro: 45.0 (#51)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Longer Query | 1483 | 1458 |
| CL-bench Life | — | 13.5% |
Writing & Preference Claude Opus 4.8 leads
Claude Opus 4.8: 72.0 (#16), DeepSeek V4 Pro: 65.5 (#46)
| Benchmark | Claude Opus 4.8 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Text | 1461 | 1451 |
| LMArena Creative Writing | 1454 | 1446 |
| EQ-Bench Creative Writing | 1840 | 1553 |
| EQ-Bench 4 | 1281 | 1166 |
| LMArena Multi-Turn | 1476 | 1467 |
Frequently asked questions
Is Claude Opus 4.8 better than DeepSeek V4 Pro?
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.3 on the Noometry Index. DeepSeek V4 Pro costs 10× 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 DeepSeek V4 Pro?
DeepSeek V4 Pro is cheaper. It lists at $0.66 per million input tokens and $1.98 per million output tokens; Claude Opus 4.8 lists at $5 and $25.
Is Claude Opus 4.8 or DeepSeek V4 Pro better for coding?
Claude Opus 4.8 scores higher on coding benchmarks: 59.9 versus 52.4 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do Claude Opus 4.8 and DeepSeek V4 Pro share?
45 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and DeepSeek V4 Pro has 48.