DeepSeek-V4-Pro-0813 vs Kimi K2.5
DeepSeek-V4-Pro-0813 (DeepSeek) and Kimi K2.5 (Moonshot) compared on benchmarks, pricing, context window, and use-case rankings.
- DeepSeek-V4-Pro-0813 leads on overall intelligence (ECI 155 vs 148).
- DeepSeek-V4-Pro-0813 is cheaper on input tokens — $0.58 vs $0.60 per 1M.
- DeepSeek-V4-Pro-0813 has a larger context window (1M vs 262K tokens).
- DeepSeek-V4-Pro-0813 ranks higher for coding (78th vs 42th percentile).
| Attribute | DeepSeek-V4-Pro-0813 | Kimi K2.5 |
|---|---|---|
| Scores | ||
| Intelligence (ECI) | 155 | 148 |
| Coding | 78 | 42 |
| Math | 88 | 49 |
| Reasoning & Knowledge | 70 | 29 |
| Agentic & Tools | — | 50 |
| Specifications | ||
| Developer | DeepSeek | Moonshot |
| Family | DeepSeek | Kimi |
| Released | Aug 13, 2026 | Feb 2, 2026 |
| Parameters | 1.6T | 1T |
| Availability | Open weights (unrestricted) | Open weights (unrestricted) |
| Context window | 1M | 262K |
| Price — $/M input | $0.58 | $0.60 |
| Price — $/M output | $1.74 | $3.00 |
| Inputs | text | text, image |
| Outputs | text | text |
| Benchmarks | ||
| AIME 2024/2025 | 99% | 92% |
| ARC-AGI | 91% | 65% |
| ARC-AGI-2 | 61% | 12% |
| GPQA Diamond | 92% | 88% |
| Humanity's Last Exam | 41% | 24% |
| SciCode | 51% | 49% |
| SimpleQA Verified | 53% | 34% |
| WebDev Arena | 1582 | 1436 |
| WeirdML | 66% | 46% |
| APEX | — | 14% |
| FrontierMath | — | 28% |
| FrontierMath Tier 4 | — | 4% |
| SimpleBench | — | 47% |
| SWE-bench Verified | — | 74% |
| Terminal-Bench | — | 43% |
| τ²-bench | — | 96% |
Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.
Frequently asked questions
Is DeepSeek-V4-Pro-0813 better than Kimi K2.5?
On Epoch AI's Capabilities Index, DeepSeek-V4-Pro-0813 scores higher (155) than Kimi K2.5 (148). The right pick depends on your task — compare their coding, math, and reasoning scores in the table above.
Which is cheaper, DeepSeek-V4-Pro-0813 or Kimi K2.5?
DeepSeek-V4-Pro-0813 is cheaper on input tokens at $0.58 per million, versus $0.60 (representative OpenRouter pricing).
Which has a larger context window, DeepSeek-V4-Pro-0813 or Kimi K2.5?
DeepSeek-V4-Pro-0813 supports up to 1M tokens, compared with 262K for the other.
Which is better for coding, DeepSeek-V4-Pro-0813 or Kimi K2.5?
Across coding benchmarks like SWE-bench Verified and Terminal-Bench, DeepSeek-V4-Pro-0813 ranks higher — 78th vs 42th percentile among the models tracked on Model Beat.
Want a different match-up? Open the compare tool to add or swap models.
More comparisons
Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.