# DeepSeek V4 Pro vs Grok 4.5

> Side-by-side benchmark comparison for DeepSeek V4 Pro and Grok 4.5 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/deepseek-v4-pro-vs-grok-4-5
- Last updated: November 21, 2026

- Shared sourced benchmarks: 15
- HTML indexing: indexable
- Ranking lane: BenchAlign v6.0

## Quick Verdict

Pick DeepSeek V4 Pro if you want the stronger benchmark profile. Grok 4.5 only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- DeepSeek V4 Pro leads overall 61.6 to 60.4.
- The clearest category separation is in reasoning, where the averages are 64.6 for DeepSeek V4 Pro and 59.9 for Grok 4.5.
- The biggest single benchmark swing is Terminal-Bench 2.1 (Vals) in Agentic, with scores of 54.7% and 67.8%.
- DeepSeek V4 Pro is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- DeepSeek V4 Pro also has the larger context window at 1M.

## Model Snapshot

| Property | DeepSeek V4 Pro | Grok 4.5 |
|----------|----------|----------|
| Creator | DeepSeek | xAI |
| Type | Open Weight | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 1M | 500K |
| Overall Score | 61.6 | 60.4 |
| Benchmarks Covered | 46 | 22 |
| Pricing (input/output) | $1.32 / $3.96 | $2.00 / $6.00 |

## Category Breakdown

### Agentic

- Winner: DeepSeek V4 Pro
- DeepSeek V4 Pro public-lane score: 60 (Supported · #55/182)
- Grok 4.5 public-lane score: 57.8 (Supported · #65/182)

| Benchmark | DeepSeek V4 Pro | Grok 4.5 | Winner |
|-----------|-----------|-----------|--------|
| Gert Labs | 52.74% | 47.24% | DeepSeek V4 Pro |
| Terminal-Bench 4.1 | 6.70% | 12.40% | Grok 4.5 |
| Finch | 67.4% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | 67.9% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | 87.9% | 83.3% | DeepSeek V4 Pro |
| BrowseComp | 83.4% | Coming soon | Coming soon |
| HLE w/ tools | 60.0% | Coming soon | Coming soon |
| MCP Atlas | 73.6% | Coming soon | Coming soon |
| Toolathlon | 51.8% | Coming soon | Coming soon |
| CyberGym | 83.3% | Coming soon | Coming soon |
| Toolathlon-Verified | 74.1% | Coming soon | Coming soon |
| Agents' Last Exam | 25.7% | Coming soon | Coming soon |
| AutomationBench | 31.8% | Coming soon | Coming soon |
| Terminal-Bench 2.1 (Vals) | 54.7% | 67.8% | Grok 4.5 |
| Terminal-Bench 3.0 | Coming soon | 15.7% | Coming soon |
| deepSwe | Coming soon | 53% | Coming soon |

### Coding

- Winner: Grok 4.5
- DeepSeek V4 Pro public-lane score: 59.6 (Supported · #56/180)
- Grok 4.5 public-lane score: 60.5 (Supported · #54/180)

| Benchmark | DeepSeek V4 Pro | Grok 4.5 | Winner |
|-----------|-----------|-----------|--------|
| LiveCodeBench Pass@1-COT | 93.5% | Coming soon | Coming soon |
| Codeforces | 3206.0 | Coming soon | Coming soon |
| SWE-bench Verified | 80.6% | Coming soon | Coming soon |
| SWE-bench Pro | 55.4% | 64.7% | Grok 4.5 |
| SWE Multilingual | 76.2% | 78% | Grok 4.5 |
| Terminal-Bench 2.0 | 67.9% | Coming soon | Coming soon |
| Vibe Code Bench | 49.93% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | 87.9% | 83.3% | DeepSeek V4 Pro |
| NL2Repo | 61.5% | Coming soon | Coming soon |
| DeepSWE | 62.7% | Coming soon | Coming soon |
| DSBench-FullStack | 71.1% | Coming soon | Coming soon |
| DSBench-Hard | 67.2% | Coming soon | Coming soon |
| OpenHarmony Bench v1.0 | 59.0% | Coming soon | Coming soon |
| LiveCodeBench (Vals) | 87.5% | 87.4% | DeepSeek V4 Pro |
| SWE-bench (Vals) | 96.4% | 86.6% | DeepSeek V4 Pro |
| CursorBench 3.2 | Coming soon | 66.7% | Coming soon |
| VulcanBench v3 | Coming soon | 89.9% | Coming soon |
| PostTrainBench v1.1 | Coming soon | 23.4% | Coming soon |
| PostTrainBench | Coming soon | 22.6% | Coming soon |

### Reasoning

- Winner: DeepSeek V4 Pro
- DeepSeek V4 Pro public-lane score: 64.6 (Supported · #32/161)
- Grok 4.5 public-lane score: 59.9 (Supported · #48/161)

| Benchmark | DeepSeek V4 Pro | Grok 4.5 | Winner |
|-----------|-----------|-----------|--------|
| MRCR 1M | 83.5% | Coming soon | Coming soon |
| CorpusQA 1M | 62.0% | Coming soon | Coming soon |
| ARC-AGI-1 | 90.00% | 85.67% | DeepSeek V4 Pro |
| ARC-AGI-2 | 61.3% | 52.6% | DeepSeek V4 Pro |
| ARC-AGI-3 | Coming soon | 0.3% | Coming soon |

### Multimodal & Grounded

- Winner: Directional only
- DeepSeek V4 Pro public-lane score: 60.7 (Estimated · #54/119)
- Grok 4.5 public-lane score: 62.9 (Supported · #43/119)

### Knowledge

- Winner: Grok 4.5
- DeepSeek V4 Pro public-lane score: 59.7 (Supported · #46/158)
- Grok 4.5 public-lane score: 61.3 (Supported · #39/158)

| Benchmark | DeepSeek V4 Pro | Grok 4.5 | Winner |
|-----------|-----------|-----------|--------|
| MMLU-Pro | 87.5% | Coming soon | Coming soon |
| Chinese-SimpleQA | 84.4% | Coming soon | Coming soon |
| GPQA | 90.1% | Coming soon | Coming soon |
| GPQA-D | 90.1% | Coming soon | Coming soon |
| HLE | 42.7% | Coming soon | Coming soon |
| GPQA Diamond (Vals) | 92.4% | 92.9% | Grok 4.5 |
| MMLU-Pro (Vals) | 87.0% | 89.2% | Grok 4.5 |

### Multilingual

- Winner: Directional only
- DeepSeek V4 Pro public-lane score: 57.3 (Estimated · #31/88)
- Grok 4.5 public-lane score: 59 (Estimated · #27/88)

### Instruction Following

- Winner: Grok 4.5
- DeepSeek V4 Pro public-lane score: 56.6 (Supported · #47/143)
- Grok 4.5 public-lane score: 57.7 (Supported · #40/143)

### Mathematics

- Winner: DeepSeek V4 Pro
- DeepSeek V4 Pro public-lane score: 59.4 (Supported · #34/125)
- Grok 4.5 public-lane score: 57.5 (Supported · #42/125)

| Benchmark | DeepSeek V4 Pro | Grok 4.5 | Winner |
|-----------|-----------|-----------|--------|
| HMMT Feb 2026 | 95.2% | Coming soon | Coming soon |
| IMOAnswerBench | 89.8% | Coming soon | Coming soon |
| Apex | 38.3% | Coming soon | Coming soon |
| Apex Shortlist | 90.2% | Coming soon | Coming soon |
| FrontierMath v2 (Tiers 1-3) | 64.561% | 57.193% | DeepSeek V4 Pro |
| FrontierMath v2 (Tier 4) | 26.829% | 24.390% | DeepSeek V4 Pro |

## FAQ

### Which is better overall, DeepSeek V4 Pro or Grok 4.5?

DeepSeek V4 Pro is ahead overall on BenchLM right now.

### Where is the biggest gap between DeepSeek V4 Pro and Grok 4.5?

The widest category gap is in reasoning, where the averages are 64.6 for DeepSeek V4 Pro and 59.9 for Grok 4.5.

### How many benchmarks does DeepSeek V4 Pro cover on BenchLM?

DeepSeek V4 Pro currently has 46 sourced benchmark scores on BenchLM.

### How many benchmarks does Grok 4.5 cover on BenchLM?

Grok 4.5 currently has 22 sourced benchmark scores on BenchLM.

## Related Comparisons

- [DeepSeek V4 Pro vs Claude Fable 5.2](/compare/claude-fable-5-2-vs-deepseek-v4-pro)
- [Grok 4.5 vs Claude Fable 5.2](/compare/claude-fable-5-2-vs-grok-4-5)
- [DeepSeek V4 Pro vs Gemini 4 Neon](/compare/deepseek-v4-pro-vs-gemini-4-neon)
- [Grok 4.5 vs Gemini 4 Neon](/compare/gemini-4-neon-vs-grok-4-5)

## Explore More

- [DeepSeek V4 Pro profile](/models/deepseek-v4-pro)
- [Grok 4.5 profile](/models/grok-4-5)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)
