Model comparison
Claude 2.1 vs DeepSeek V4 Pro
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 25.2 on the Noometry Index.
Last verified . 6 shared benchmarks.
DeepSeekSummary
- They share 6 benchmarks with published results for both. Claude 2.1 scores higher in 0 categories and DeepSeek V4 Pro in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Pro leads 64.8 to 10.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.9% for Claude 2.1 and 98.6% for DeepSeek V4 Pro.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| Claude 2.1 | DeepSeek V4 Pro | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 25.2 | 54.3 |
| Released | 2023-11-21 | 2026-04-24 |
| Weights | Proprietary | Open |
| Context window | — | 1M |
| Max output | — | 393K |
| Input $ / M tokens | — | $0.66 |
| Output $ / M tokens | — | $1.98 |
| Results tracked | 7 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
Claude 2.1: 26.2 (#327), DeepSeek V4 Pro: 52.4 (#34)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| WeirdML | 7.1% | 66.2% |
| SWE-bench Verified | — | 77.6% |
| FrontierCode | — | 28.6% |
| LMArena WebDev | — | 1582 |
| SciCode | — | 51% |
| LMArena Coding | — | 1470 |
| ALE-Bench | — | 1,403 |
Agentic & Tool Use Not comparable
Claude 2.1: —, DeepSeek V4 Pro: 32.8 (#58)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| APEX-Agents | — | 47.3% |
| Vending-Bench 2 | — | 3,285 |
Reasoning DeepSeek V4 Pro leads
Claude 2.1: 21.4 (#221), DeepSeek V4 Pro: 56.5 (#24)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| DTBench | 51% | 93.9% |
| Epoch Capabilities Index | 119.27 | 155.31 |
| ForecastBench | 54.2 | 56.1 |
| ARC-AGI-2 | — | 61.3% |
| Kagi LLM Benchmark | — | 53.5% |
| NYT Connections (extended) | — | 91.3% |
| ARC-AGI-1 | — | 90.5% |
| CritPt | — | 18% |
| Chess Puzzles | — | 47% |
| LMArena Hard Prompts | — | 1461 |
| Mystery Game Puzzles | — | 43% |
| LMCA | — | 45.5% |
| Surface Evolver Bench | — | 40% |
Math DeepSeek V4 Pro leads
Claude 2.1: 10.2 (#315), DeepSeek V4 Pro: 64.8 (#30)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.9% | 98.6% |
| FrontierMath (Tiers 1-3) | — | 64.6% |
| FrontierMath Tier 4 | — | 26.8% |
| MathArena Final-Answer Competitions | — | 76.6% |
| ProofBench | — | 50% |
| LMArena Math | — | 1455 |
Knowledge DeepSeek V4 Pro leads
Claude 2.1: 15.4 (#292), DeepSeek V4 Pro: 59.5 (#31)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| GPQA Diamond | 33% | 91.7% |
| SimpleQA Verified | — | 52.9% |
| Vectara Hallucination Rate | — | 8.6% |
| LMArena Expert | — | 1464 |
| MMLU | 73.5% | — |
Multilingual Not comparable
Claude 2.1: —, DeepSeek V4 Pro: 54.4 (#45)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Non-English | — | 1439 |
| LMArena Chinese | — | 1486 |
| LMArena French | — | 1472 |
| LMArena German | — | 1458 |
| LMArena Japanese | — | 1445 |
| LMArena Korean | — | 1447 |
| LMArena Russian | — | 1453 |
| LMArena Spanish | — | 1458 |
Instruction Following Not comparable
Claude 2.1: —, DeepSeek V4 Pro: 76.1 (#47)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Instruction Following | — | 1448 |
Long Context Not comparable
Claude 2.1: —, DeepSeek V4 Pro: 45.0 (#51)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| CL-bench Life | — | 13.5% |
| LMArena Longer Query | — | 1458 |
Writing & Preference Not comparable
Claude 2.1: —, DeepSeek V4 Pro: 65.5 (#46)
| Benchmark | Claude 2.1 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Text | — | 1451 |
| LMArena Creative Writing | — | 1446 |
| EQ-Bench Creative Writing | — | 1553 |
| EQ-Bench 4 | — | 1166 |
| LMArena Multi-Turn | — | 1467 |
Frequently asked questions
Is Claude 2.1 better than DeepSeek V4 Pro?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 25.2 on the Noometry Index.
Is Claude 2.1 or DeepSeek V4 Pro better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 26.2 in the Noometry coding category.
How many benchmarks do Claude 2.1 and DeepSeek V4 Pro share?
6 benchmarks have published results for both models. Claude 2.1 has 7 scored results on Noometry and DeepSeek V4 Pro has 48.