# DeepSeek V4 Pro: price, benchmarks, context

- **URL canônica:** https://meirinho.com.br/en/ia/modelos/deepseek-v4-pro/
- **Idioma:** en
- **Atualizado:** 2026-11-17

---
**19th–44th of 85**

- technical tie with the top score — OTIS Mock AIME

Methodology v2 · updated November 21, 2026 · [Technical tie](https://meirinho.com.br/en/ia/modelos/metodologia/#empate-tecnico) · [Minimum of 4 benchmarks](https://meirinho.com.br/en/ia/modelos/metodologia/#minimo-3-benchmarks) · [Methodology](https://meirinho.com.br/en/ia/modelos/metodologia/)

DeepSeek

# DeepSeek V4 Pro

Benchmarks, price, context and versions

GPQA Diamond

Epoch AI

FrontierMath (tiers 1 to 3)

Epoch AI

Input US$/1M

Official pricing

Output US$/1M

Official pricing

Context

tokens

## Direct answer

**Direct answer:** DeepSeek V4 Pro is a model from DeepSeek. 90.9% on GPQA Diamond, per Epoch AI. US$ 1.32 per 1 million input tokens and US$ 3.96 per 1 million output tokens, peak hours.

## ✓ Good for

- technical reasoning on science topics (35th of 195 on GPQA Diamond, in a statistical tie with 3 other models, per Epoch AI)
- fixing bugs in real code (6th of 31 on SWE-bench Verified, per Epoch AI)
- technical reasoning as the model appears through the API (22nd of 134 on GPQA Diamond (OpenRouter), in a statistical tie with 2 other models, per OpenRouter Benchmarks)
- very long documents or code (a window of 1 million tokens or more, according to OpenRouter)
- running on your own infrastructure (open weights)

## ✕ Not the best for

- the ceiling of reasoning on research-level math (56th of 85 on FrontierMath (tiers 1 to 3), per Epoch AI)

This block follows a rule of this section, using source data: if the model is in the top 25% of the metric, the item goes under "Good for"; if it is in the bottom half, it goes under "Not the best for". The middle band appears in neither. The price is official pricing, published by the lab itself for its own API. Context uses OpenRouter. Scores use Epoch AI or OpenRouter Benchmarks, in the latest version we checked.

## Benchmarks

| GPQA Diamond[? what it is](https://meirinho.com.br/en/ia/modelos/glossario/gpqa-diamond/) | 90.9%**highest score on this metric 96.8%35º / 195 · [Epoch AI](https://epoch.ai/benchmarks/gpqa-diamond), latest version we checked |
| --- | --- |
| OTIS Mock AIME[?](https://meirinho.com.br/en/ia/modelos/glossario/otis-mock-aime/) | 96.7%**highest score on this metric 100%38º / 182 · [Epoch AI](https://epoch.ai/benchmarks/otis-mock-aime-2024-2025), latest version we checked |
| FrontierMath (tiers 1 to 3)[?](https://meirinho.com.br/en/ia/modelos/glossario/frontiermath/) | 45.3%**highest score on this metric 95.1%56º / 85 · [Epoch AI](https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2), latest version we checked |
| SimpleQA Verified[?](https://meirinho.com.br/en/ia/modelos/glossario/simpleqa-verified/) | 47%**highest score on this metric 94%38º / 87 · [Epoch AI](https://epoch.ai/benchmarks/simpleqa-verified), latest version we checked |
| SWE-bench Verified[?](https://meirinho.com.br/en/ia/modelos/glossario/swe-bench-verified/) | 77.6%**highest score on this metric 83.5%6º / 31 · [Epoch AI](https://epoch.ai/benchmarks/swe-bench-verified), latest version we checked |
| GPQA Diamond (OpenRouter)[?](https://meirinho.com.br/en/ia/modelos/glossario/gpqa-diamond-openrouter/) | 91.9%**highest score on this metric 96.5%22º / 134 · [OpenRouter Benchmarks](https://openrouter.ai/benchmarks/gpqa-diamond), latest version we checked |
| τ-bench Verified (airline)[?](https://meirinho.com.br/en/ia/modelos/glossario/tau-bench/) | 78.6%**highest score on this metric 82%16º / 130 · [OpenRouter Benchmarks](https://openrouter.ai/benchmarks/tau2-bench-airline), latest version we checked |

## Input US$/1M · Output US$/1M

Input US$/1M

US$ 1.32

Third-party provider pricing, OpenRouter: US$ 0.90 / US$ 1.50

- off-peak hours: US$ 0.66 / US$ 1.98 · Cache read US$/1M US$ 0.022

Output US$/1M

US$ 3.96

Cache read US$/1M

US$ 0.044

## Key specs

| Lab | [DeepSeek](https://meirinho.com.br/en/ia/modelos/lab/deepseek/) |
| --- | --- |
| Origin | China |
| Family | DeepSeek V4 provisional data |
| Weights[?](https://meirinho.com.br/en/ia/modelos/glossario/pesos-abertos/) | open |
| License | no data |
| Also known as | API name: [deepseek-v4-pro](https://api-docs.deepseek.com/quick_start/pricing/#model-details) |
| Context[?](https://meirinho.com.br/en/ia/modelos/glossario/janela-de-contexto/) | 1,048,576 tokens |
| Parameters | no data |
| Input | text |
| OpenRouter usage[?](https://meirinho.com.br/en/ia/modelos/glossario/uso-no-openrouter/) | 0.61% OpenRouter |
| Downloads (30 days)[?](https://meirinho.com.br/en/ia/modelos/glossario/downloads-hugging-face/) | no data |
| Use cases provisional data | [Text](https://meirinho.com.br/en/ia/modelos/uso/texto/) [Agents](https://meirinho.com.br/en/ia/modelos/uso/agentes/) [Code](https://meirinho.com.br/en/ia/modelos/uso/codigo/) |
| Dates in sources | no official dateApril 24, 2026, published, per Epoch AIApril 24, 2026, listed on OpenRouter |

Collected November 17, 2026

## Version history

- August 13, 2026 [DeepSeek-V4-Pro-0813](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813) Hugging Face Hub, Hugging Face repo created
- August 12, 2026 [DeepSeek: DeepSeek V4 Pro 0813](https://openrouter.ai/deepseek/deepseek-v4-pro-0813) OpenRouter, listed on OpenRouter
- April 24, 2026 [DeepSeek: DeepSeek V4 Pro 0423](https://openrouter.ai/deepseek/deepseek-v4-pro) OpenRouter, listed on OpenRouter

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch.

## Sources for this model

- Official model page, [DeepSeek](https://api-docs.deepseek.com/quick_start/pricing/).
- Official pricing, [DeepSeek](https://api-docs.deepseek.com/quick_start/pricing/#model-details).
- No official launch announcement.
- Benchmarks: each score carries its own source, in the table above.

## FAQ

US$ 1.32 per 1 million input tokens and US$ 3.96 per 1 million output tokens, peak hours.

Yes, it has open weights. License: no data.

90.9% on GPQA Diamond, per Epoch AI.

DeepSeek V4 Pro is a model from DeepSeek, with open weights, used for text, agents, code.

1,048,576 tokens, per OpenRouter.

It scores 77.6% on SWE-bench Verified, per Epoch AI.

It was made by DeepSeek, a lab from China.

provisional data Questions ordered by what shows up in public Google search suggestions (no volume). Answers use source data only.

## Compare with

### Compare with

- [GPT-6.1 Sol vs DeepSeek V4 Pro](https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v4-pro-vs-gpt-6-1-sol/)
- [DeepSeek V4 Pro vs Qwen3.8 Max](https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v4-pro-vs-qwen3-8-max/)
- [DeepSeek V4 Pro vs Kimi K3.1 Pro](https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v4-pro-vs-kimi-k3-1-pro/)
- [DeepSeek V4 Pro vs GLM-5.3](https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v4-pro-vs-glm-5-3/)

### Models from DeepSeek

- [DeepSeek V4.1 Flash](https://meirinho.com.br/en/ia/modelos/deepseek-v4-1-flash/)
- [DeepSeek V4 Flash 0423](https://meirinho.com.br/en/ia/modelos/deepseek-v4-flash/)
- [DeepSeek V3.2](https://meirinho.com.br/en/ia/modelos/deepseek-v3-2/)
- [DeepSeek V4 Flash Vision Exp](https://meirinho.com.br/en/ia/modelos/deepseek-v4-flash-vision-exp/)
- [DeepSeek V3.1](https://meirinho.com.br/en/ia/modelos/deepseek-chat-v3-1/)
- [DeepSeek V3](https://meirinho.com.br/en/ia/modelos/deepseek-v3/)
- [DeepSeek V3.2 Exp](https://meirinho.com.br/en/ia/modelos/deepseek-v3-2-exp/)
- [DeepSeek R1](https://meirinho.com.br/en/ia/modelos/deepseek-r1/)
- [DeepSeek V3.1 Terminus](https://meirinho.com.br/en/ia/modelos/deepseek-v3-1-terminus/)
- [DeepSeek LLM 67B](https://meirinho.com.br/en/ia/modelos/deepseek-llm-67b/)
- [DeepSeek V4 Flash 0731](https://meirinho.com.br/en/ia/modelos/deepseek-v4-flash-0731/)
- [DeepSeek-R1-Distill-Llama-70B](https://meirinho.com.br/en/ia/modelos/deepseek-r1-distill-llama-70b/)

[See lab](https://meirinho.com.br/en/ia/modelos/lab/deepseek/)

## How to cite

If you use this data, cite the page and the date you accessed it.

```
Adriano Meirinho. "DeepSeek V4 Pro". meirinho.com.br. https://meirinho.com.br/en/ia/modelos/deepseek-v4-pro/. Accessed November 21, 2026.
```

Third-party data follows the license of each source. The full table is available as [CSV](https://meirinho.com.br/ia/modelos/dados.csv) and [JSON](https://meirinho.com.br/ia/modelos/dados.json).

[#](#)
