# GPT-5 vs DeepSeek V3: benchmarks and price

- **URL canônica:** https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v3-vs-gpt-5/
- **Idioma:** en
- **Atualizado:** 2026-11-17

---
AI models

# GPT-5 vs DeepSeek V3

Same criteria, side by side. Where the gap is small, it may be measurement noise.

GPT-5

Input US$/1M / Output US$/1M · Official pricing

DeepSeek V3

Input US$/1M / Output US$/1M · Third-party provider

GPQA Diamond

86.2% · 67.6%

GPT-5 · DeepSeek V3

## Direct answer

Pick GPT-5 if scores come first: it leads on GPQA Diamond and τ-bench Verified. Pick DeepSeek V3 if cost matters more: it charges US$ 0.29 versus US$ 1.25 per 1M input tokens and US$ 1.14 versus US$ 10.00 per 1M output.

GPT-5

- GPQA Diamond **+18.6 p.p.**
- τ-bench Verified **+26.9 p.p.**

statistical tie

| Metric | GPT-5 | DeepSeek V3 | Source |
| --- | --- | --- | --- |
| Input US$/1M | US$ 1.25 / US$ 10.00 Official pricing | US$ 0.29 / US$ 1.14 Third-party provider |  |
| Output US$/1M | US$ 10.00 | US$ 1.14 |  |
| Context | 400,000 | 163,840 |  |
| BENCHMARKS |
| GPQA Diamond | **86.2%****between 83rd and 96th of 250 | 67.6%**between 149th and 153rd of 250 | [Epoch AI](https://epoch.ai/benchmarks/gpqa-diamond) |
| OTIS Mock AIME saturated | 91.4%** statistical tie with the top score | 37.8%** | [Epoch AI](https://epoch.ai/benchmarks/otis-mock-aime-2024-2025) |
| τ-bench Verified (airline) | **70.7%****between 80th and 89th of 130 | 43.8%**between 119th and 122nd of 130 | [OpenRouter Benchmarks](https://openrouter.ai/benchmarks/tau2-bench-airline) |
| OTIS Mock AIME saturated | 91.4%** statistical tie with the top score | 37.8%** | [Epoch AI](https://epoch.ai/benchmarks/otis-mock-aime-2024-2025) |

## Compare with

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

## Sources

- [Epoch AI](https://epoch.ai/benchmarks), CC BY 4.0, latest version we checked.
- [OpenRouter Benchmarks](https://openrouter.ai/benchmarks), CC BY 4.0, latest version we checked.
- [OpenRouter](https://openrouter.ai/api/v1/models), dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.

## How to cite

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

```
Adriano Meirinho. "GPT-5 vs DeepSeek V3". meirinho.com.br. https://meirinho.com.br/en/ia/modelos/comparar/deepseek-v3-vs-gpt-5/. 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).

[#](#)
