# DeepSeek V4 Pro vs Llama 4 Maverick

> DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 30.9 on the Noometry Index. Llama 4 Maverick costs 3.3× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v4-pro-vs-llama-4-maverick
- Last updated: 2026-10-10
- Shared benchmarks: 33

## Summary

- They share 33 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and Llama 4 Maverick in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 10.1.
- The biggest single-benchmark swing is ARC-AGI-1: 90.5% for DeepSeek V4 Pro and 4.4% for Llama 4 Maverick.
- Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 128K.

## Snapshot

| | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 54.3 | 30.9 |
| Rank | 31 | 282 |
| Context | 1M | 128K |
| Input $/M | $0.66 | $0.19 |
| Output $/M | $1.98 | $0.65 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Pro: 52.4 (#34)
- Llama 4 Maverick: 26.6 (#324)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| SciCode | 51% | 33.1% |
| WeirdML | 66.2% | 24.5% |
| LMArena Coding | 1470 | 1302 |
| ALE-Bench | 1,403 | 172.97 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| SWE-bench Verified (bash only) | — | 21% |
| Aider Polyglot | — | 15.6% |
| LMArena WebDev | 1582 | — |
| BigCodeBench Instruct | — | 49.7% |
| BigCodeBench Complete | — | 61.4% |

## Agentic & Tool Use

- DeepSeek V4 Pro: 32.8 (#58)
- Llama 4 Maverick: 28.2 (#91)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Berkeley Function Calling Leaderboard | — | 37.3% |
| Vending-Bench 2 | 3,285 | — |

## Reasoning

- DeepSeek V4 Pro: 56.5 (#24)
- Llama 4 Maverick: 10.1 (#342)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 61.3% | 0% |
| Kagi LLM Benchmark | 53.5% | 55.9% |
| NYT Connections (extended) | 91.3% | 8% |
| ARC-AGI-1 | 90.5% | 4.4% |
| CritPt | 18% | 0% |
| LMArena Hard Prompts | 1461 | 1281 |
| DTBench | 93.9% | 61.9% |
| LMCA | 45.5% | 15.9% |
| Epoch Capabilities Index | 155.31 | 132.2 |
| ForecastBench | 56.1 | 57.5 |
| SimpleBench | — | 27.7% |
| Chess Puzzles | 47% | — |
| EnigmaEval | — | 0.6% |
| Mystery Game Puzzles | 43% | — |
| Surface Evolver Bench | 40% | — |

## Math

- DeepSeek V4 Pro: 64.8 (#30)
- Llama 4 Maverick: 26.0 (#262)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 20.6% |
| LMArena Math | 1455 | 1299 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
| Omni-MATH | — | 42.2% |
| MATH Level 5 | — | 73% |
| FrontierMath (Feb 2025 set) | — | 0.7% |

## Knowledge

- DeepSeek V4 Pro: 59.5 (#31)
- Llama 4 Maverick: 33.4 (#204)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 91.7% | 67% |
| Vectara Hallucination Rate | 8.6% | 8.2% |
| LMArena Expert | 1464 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| SimpleQA Verified | 52.9% | — |
| MMLU-Pro | — | 81% |
| Confabulations | — | 22.6% |
| GPQA (HELM) | — | 65% |

## Multimodal

- DeepSeek V4 Pro: —
- Llama 4 Maverick: 31.6 (#105)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | — | 1142 |
| GeoBench | — | 52% |
| SpatialViz-Bench | — | 31.8% |

## Multilingual

- DeepSeek V4 Pro: 54.4 (#45)
- Llama 4 Maverick: 42.2 (#195)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1439 | 1269 |
| LMArena Chinese | 1486 | 1277 |
| LMArena French | 1472 | 1259 |
| LMArena German | 1458 | 1291 |
| LMArena Japanese | 1445 | 1207 |
| LMArena Korean | 1447 | 1203 |
| LMArena Russian | 1453 | 1286 |
| LMArena Spanish | 1458 | 1293 |

## Instruction Following

- DeepSeek V4 Pro: 76.1 (#47)
- Llama 4 Maverick: 71.7 (#146)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Instruction Following | 1448 | 1267 |
| IFEval | — | 90.8% |

## Long Context

- DeepSeek V4 Pro: 45.0 (#51)
- Llama 4 Maverick: 31.4 (#279)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1458 | 1280 |
| Fiction.LiveBench | — | 46.2% |
| CL-bench Life | 13.5% | — |

## Writing & Preference

- DeepSeek V4 Pro: 65.5 (#46)
- Llama 4 Maverick: 38.8 (#252)

| Benchmark | DeepSeek V4 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1451 | 1287 |
| LMArena Creative Writing | 1446 | 1267 |
| EQ-Bench Creative Writing | 1553 | 860 |
| LMArena Multi-Turn | 1467 | 1289 |
| Short-Story Creative Writing | — | 62% |
| WildBench | — | 80% |
| EQ-Bench 4 | 1166 | — |

## FAQ

### Is DeepSeek V4 Pro better than Llama 4 Maverick?

DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 30.9 on the Noometry Index. Llama 4 Maverick costs 3.3× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek V4 Pro or Llama 4 Maverick?

Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.

### Is DeepSeek V4 Pro or Llama 4 Maverick better for coding?

DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 26.6 in the Noometry coding category.

### Which has the bigger context window?

DeepSeek V4 Pro does, with 1M tokens against 128K.

### How many benchmarks do DeepSeek V4 Pro and Llama 4 Maverick share?

33 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Llama 4 Maverick has 54.