Model comparison
DeepSeek V4 Pro vs Gemini 4 Argon
Gemini 4 Argon is the stronger model overall, scoring 56.5 to 54.3 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeek
GoogleSummary
- They share 19 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 2 categories and Gemini 4 Argon in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Gemini 4 Argon leads 49.2 to 32.8.
- The biggest single-benchmark swing is ProofBench: 50% for DeepSeek V4 Pro and 99% for Gemini 4 Argon.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | Gemini 4 Argon | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 54.3 | 56.5 |
| Released | 2026-04-24 | 2026-09-30 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 4 Argon leads
DeepSeek V4 Pro: 52.4 (#34), Gemini 4 Argon: 62.0 (#9)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena WebDev | 1582 | 1678 |
| SciCode | 51% | 61.8% |
| LMArena Coding | 1470 | 1550 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| FrontierSWE | — | 55% |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Gemini 4 Argon leads
DeepSeek V4 Pro: 32.8 (#58), Gemini 4 Argon: 49.2 (#8)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| APEX-Agents | 47.3% | 82.2% |
| Vending-Bench 2 | 3,285 | 13,718 |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Gemini 4 Argon: 44.2 (#47)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| CritPt | 18% | 27.1% |
| LMArena Hard Prompts | 1461 | 1548 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
| ForecastBench | 56.1 | — |
Math Gemini 4 Argon leads
DeepSeek V4 Pro: 64.8 (#30), Gemini 4 Argon: 73.9 (#17)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| ProofBench | 50% | 99% |
| LMArena Math | 1455 | 1536 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Gemini 4 Argon: 43.9 (#89)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena Expert | 1464 | 1553 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
Multimodal Not comparable
DeepSeek V4 Pro: —, Gemini 4 Argon: 46.5 (#14)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| Blueprint-Bench 2 | — | 54.4% |
Multilingual Gemini 4 Argon leads
DeepSeek V4 Pro: 54.4 (#45), Gemini 4 Argon: 60.3 (#1)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena Non-English | 1439 | 1523 |
| LMArena Chinese | 1486 | 1610 |
| LMArena Russian | 1453 | 1537 |
| LMArena Spanish | 1458 | 1497 |
| LMArena French | 1472 | — |
| LMArena German | 1458 | — |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
Instruction Following Gemini 4 Argon leads
DeepSeek V4 Pro: 76.1 (#47), Gemini 4 Argon: 80.1 (#2)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena Instruction Following | 1448 | 1538 |
Long Context Gemini 4 Argon leads
DeepSeek V4 Pro: 45.0 (#51), Gemini 4 Argon: 47.6 (#15)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena Longer Query | 1458 | 1549 |
| CL-bench Life | 13.5% | — |
Writing & Preference Gemini 4 Argon leads
DeepSeek V4 Pro: 65.5 (#46), Gemini 4 Argon: 71.4 (#19)
| Benchmark | DeepSeek V4 Pro | Gemini 4 Argon |
|---|---|---|
| LMArena Text | 1451 | 1534 |
| LMArena Creative Writing | 1446 | 1531 |
| LMArena Multi-Turn | 1467 | 1557 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Gemini 4 Argon?
Gemini 4 Argon is the stronger model overall, scoring 56.5 to 54.3 on the Noometry Index.
Is DeepSeek V4 Pro or Gemini 4 Argon better for coding?
Gemini 4 Argon scores higher on coding benchmarks: 62.0 versus 52.4 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and Gemini 4 Argon share?
19 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Gemini 4 Argon has 21.