# DeepSeek V4 Pro 0813 vs Kimi K2.5

> Side-by-side benchmark comparison for DeepSeek V4 Pro 0813 and Kimi K2.5 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/deepseek-v4-pro-0813-vs-kimi-k2-5
- Last updated: September 27, 2026

- Shared sourced benchmarks: 12
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.7

## Quick Verdict

Pick DeepSeek V4 Pro 0813 if you want the stronger benchmark profile. Kimi K2.5 only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- DeepSeek V4 Pro 0813 leads overall 63.48 to 52.96.
- The clearest category separation is in agentic, where the averages are 55 for DeepSeek V4 Pro 0813 and 36.1 for Kimi K2.5.
- The biggest single benchmark swing is MCP Atlas in Agentic, with scores of 73.6% and 29.5%.
- Kimi K2.5 is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- DeepSeek V4 Pro 0813 also has the larger context window at 1M.

## Model Snapshot

| Property | DeepSeek V4 Pro 0813 | Kimi K2.5 |
|----------|----------|----------|
| Creator | DeepSeek | Moonshot AI |
| Type | Open Weight | Open Weight |
| Reasoning | Reasoning | Non-Reasoning |
| Context | 1M | 256K |
| Overall Score | 63.48 | 52.96 |
| Benchmarks Covered | 42 | 45 |
| Pricing (input/output) | $1.32 / $3.96 | $0.60 / $3.00 |

## Category Breakdown

### Agentic

- Winner: Directional only
- DeepSeek V4 Pro 0813 public-lane score: 55 (Supported · #29/105)
- Kimi K2.5 public-lane score: 36.1 (Estimated · #52/105)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.0 | 67.9% | 50.8% | DeepSeek V4 Pro 0813 |
| Terminal-Bench 2.1 | 87.9% | Coming soon | Coming soon |
| BrowseComp | 83.4% | 60.6% | DeepSeek V4 Pro 0813 |
| HLE w/ tools | 60.0% | Coming soon | Coming soon |
| MCP Atlas | 73.6% | 29.5% | DeepSeek V4 Pro 0813 |
| Toolathlon | 51.8% | 27.8% | DeepSeek V4 Pro 0813 |
| 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% | Coming soon | Coming soon |
| Claw-Eval | Coming soon | 52.3% | Coming soon |
| QwenClawBench | Coming soon | 54.3% | Coming soon |
| τ³-bench results | Coming soon | 65.7% | Coming soon |
| DeepSearchQA | Coming soon | 77.1% | Coming soon |
| DeepPlanning | Coming soon | 14.4% | Coming soon |
| MCP-Tasks | Coming soon | 59.1% | Coming soon |
| WideResearch | Coming soon | 72.7% | Coming soon |
| Gert Labs | Coming soon | 45.88% | Coming soon |
| ResearchClawBench | Coming soon | 14.0% | Coming soon |
| JobBench | Coming soon | 8.7% | Coming soon |

### Coding

- Winner: Directional only
- DeepSeek V4 Pro 0813 public-lane score: 50.3 (Supported · #39/135)
- Kimi K2.5 public-lane score: 38.9 (Estimated · #61/135)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| LiveCodeBench Pass@1-COT | 93.5% | Coming soon | Coming soon |
| Codeforces | 3206.0 | Coming soon | Coming soon |
| SWE-bench Verified | 80.6% | 76.8% | DeepSeek V4 Pro 0813 |
| SWE-bench Pro | 55.4% | 50.7% | DeepSeek V4 Pro 0813 |
| SWE Multilingual | 76.2% | 73% | DeepSeek V4 Pro 0813 |
| 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% | Coming soon | Coming soon |
| 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 | 59.0% | Coming soon | Coming soon |
| LiveCodeBench (Vals) | 87.5% | Coming soon | Coming soon |
| SWE-bench (Vals) | 96.4% | Coming soon | Coming soon |
| SWE-bench Verified* | Coming soon | 70.8% | Coming soon |
| LiveCodeBench v6 | Coming soon | 85.0% | Coming soon |
| SWE-Rebench | Coming soon | 58.5% | Coming soon |
| React Native Evals | Coming soon | 77.2% | Coming soon |
| SciCode | Coming soon | 48.7% | Coming soon |

### Reasoning

- Winner: Not comparable
- DeepSeek V4 Pro 0813 public-lane score: 56.9 (Unranked · 4 rankable rows)
- Kimi K2.5 public-lane score: 55.1 (Unranked · 3 rankable rows)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| MRCR 1M | 83.5% | Coming soon | Coming soon |
| CorpusQA 1M | 62.0% | Coming soon | Coming soon |
| ARC-AGI-1 | 90.00% | Coming soon | Coming soon |
| ARC-AGI-2 | 61.3% | Coming soon | Coming soon |
| LongBench v2 | Coming soon | 61% | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- DeepSeek V4 Pro 0813 public-lane score: Coming soon
- Kimi K2.5 public-lane score: 65.8 (#25/50)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | Coming soon | 78.5% | Coming soon |
| Video-MME | Coming soon | 87.4% | Coming soon |
| MMVU | Coming soon | 80.4% | Coming soon |
| VideoMMMU | Coming soon | 86.6% | Coming soon |

### Knowledge

- Winner: Directional only
- DeepSeek V4 Pro 0813 public-lane score: 63.4 (Estimated · #29/158)
- Kimi K2.5 public-lane score: 47.8 (Estimated · #65/158)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| MMLU-Pro | 87.5% | 87.1% | DeepSeek V4 Pro 0813 |
| SimpleQA | 57.9% | Coming soon | Coming soon |
| Chinese-SimpleQA | 84.4% | Coming soon | Coming soon |
| GPQA | 90.1% | 87.6% | DeepSeek V4 Pro 0813 |
| GPQA-D | 90.1% | 87.6% | DeepSeek V4 Pro 0813 |
| HLE | 42.7% | 30.1% | DeepSeek V4 Pro 0813 |
| GPQA Diamond (Vals) | 92.4% | Coming soon | Coming soon |
| MMLU-Pro (Vals) | 87.0% | Coming soon | Coming soon |
| SuperGPQA | Coming soon | 69.2% | Coming soon |
| MMLU-Pro (Arcee) | Coming soon | 87.1% | Coming soon |

### Multilingual

- Winner: Not comparable
- DeepSeek V4 Pro 0813 public-lane score: Coming soon
- Kimi K2.5 public-lane score: 38.2 (#8/12)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| MMLU-ProX | Coming soon | 82.3% | Coming soon |
| NOVA-63 | Coming soon | 56.0% | Coming soon |

### Instruction Following

- Winner: Not comparable
- DeepSeek V4 Pro 0813 public-lane score: Coming soon
- Kimi K2.5 public-lane score: 84.4 (#41/124)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| IFEval | Coming soon | 93.9% | Coming soon |

### Mathematics

- Winner: Not comparable
- DeepSeek V4 Pro 0813 public-lane score: 80.2 (Unranked · 4 rankable rows)
- Kimi K2.5 public-lane score: 62.1 (#4/7)

| Benchmark | DeepSeek V4 Pro 0813 | Kimi K2.5 | Winner |
|-----------|-----------|-----------|--------|
| HMMT Feb 2026 | 95.2% | 87.1% | DeepSeek V4 Pro 0813 |
| IMOAnswerBench | 89.8% | Coming soon | Coming soon |
| Apex | 38.3% | Coming soon | Coming soon |
| Apex Shortlist | 90.2% | Coming soon | Coming soon |
| AIME 2025 | Coming soon | 96.1% | Coming soon |
| AIME26 | Coming soon | 95.8% | Coming soon |
| AIME25 (Arcee) | Coming soon | 96.3% | Coming soon |
| HMMT Feb 2025 | Coming soon | 95.4% | Coming soon |
| HMMT Nov 2025 | Coming soon | 91.1% | Coming soon |
| MMAnswerBench | Coming soon | 81.8% | Coming soon |
| FrontierMath v2 (Tiers 1-3) | Coming soon | 27.900% | Coming soon |
| FrontierMath v2 (Tier 4) | Coming soon | 4.200% | Coming soon |

## FAQ

### Which is better overall, DeepSeek V4 Pro 0813 or Kimi K2.5?

DeepSeek V4 Pro 0813 is ahead overall on BenchLM right now.

### Where is the biggest gap between DeepSeek V4 Pro 0813 and Kimi K2.5?

The widest category gap is in agentic, where the averages are 55 for DeepSeek V4 Pro 0813 and 36.1 for Kimi K2.5.

### How many benchmarks does DeepSeek V4 Pro 0813 cover on BenchLM?

DeepSeek V4 Pro 0813 currently has 42 sourced benchmark scores on BenchLM.

### How many benchmarks does Kimi K2.5 cover on BenchLM?

Kimi K2.5 currently has 45 sourced benchmark scores on BenchLM.

## Related Comparisons

- [DeepSeek V4 Pro 0813 vs Kimi K2.5 (Reasoning)](/compare/deepseek-v4-pro-0813-vs-kimi-k2-5-reasoning)
- [Kimi K2.5 vs Kimi K2.5 (Reasoning)](/compare/kimi-k2-5-vs-kimi-k2-5-reasoning)
- [DeepSeek V4 Pro 0813 vs GPT-6 Astra](/compare/deepseek-v4-pro-0813-vs-gpt-6-astra)
- [Kimi K2.5 vs GPT-6 Astra](/compare/gpt-6-astra-vs-kimi-k2-5)

## Explore More

- [DeepSeek V4 Pro 0813 profile](/models/deepseek-v4-pro-0813)
- [Kimi K2.5 profile](/models/kimi-k2-5)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)