Deepseek | L’Alternative Ultime à ChatGPT ? (100% gratuite)

Deepseek | L’Alternative Ultime à ChatGPT ? (100% gratuite)

by Shubham SHARMA

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Shubham SHARMA 👥 313K 📅 January 30, 2025 ⏱ 12 min 👁 63K 📄 tutorial 🧭 2026-08-24
Available in: English (current) Français

Keywords

DeepSeek R1OllamaOpen SourceLocal deploymentData privacy

Summary

The video begins with a rapid recap of DeepSeek’s emergence as a Chinese AI company offering open-source models, contrasting with proprietary models from OpenAI, Google, and Anthropic. It highlights DeepSeek R1’s performance comparable to OpenAI’s O1 at a fraction of the training cost (5.5 million vs 100 million dollars), and its rapid rise to the top of the App Store. The presenter then addresses the ‘free product’ concern by explaining that using DeepSeek’s hosted service means your data goes to their servers, but because the model is open-source, you can run it locally. He demonstrates installing DeepSeek R1 via Ollama on a personal computer, using a distilled version (1.5B parameters) for practicality, and then shows how to set up a user-friendly interface with Open WebUI. For automation, he explains the API cost difference (DeepSeek is about 2% of O1’s price) and shows how to deploy Ollama on a Hostinger VPS to have a private, secure AI instance accessible via API. He also mentions using n8n for open-source automation and hints at advanced tools like Bolt.diy and Browser Use for building applications and automating web tasks with local models. The video concludes by inviting viewers to request more in-depth tutorials on open-source tools.

202 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable practical information for users interested in data privacy and cost savings by running open-source AI models locally. The argumentation is clear and persuasive, using analogies (e.g., training a dog) to explain complex concepts like model training and hosting. The step-by-step tutorial is easy to follow, and the emphasis on self-hosting addresses common privacy concerns. However, the argumentation is somewhat one-sided, promoting the benefits of open-source without deeply discussing limitations such as model quality of distilled versions or the technical expertise required for server setup.

97 words

Title / Content Match

The title accurately reflects the content, which focuses on DeepSeek as a free alternative to ChatGPT, with practical installation steps.

Quality & Reliability

6/10

The video provides a practical tutorial on installing and using DeepSeek R1 locally via Ollama and WebUI, with a clear explanation of open-source benefits and data privacy. However, it lacks rigorous scientific depth, contains promotional content for Hostinger, and makes claims about training costs and performance without citing primary sources.

Chapters

Cited Sources

  • Ollama — Official website for Ollama, the tool used to run local AI models.
  • Open WebUI — Official website for Open WebUI, the interface used to interact with local models.
  • Hostinger VPS — Hostinger VPS service, used in the video to deploy a private AI server.

Concurring Sources

  • Ollama — Official tool for running local models, consistent with the video's tutorial.
  • Open WebUI — Interface used in the video, consistent with its purpose.

Dissenting Sources

  • OpenAI — OpenAI's proprietary models are the main competitors to DeepSeek, and the video's claims about cost and performance are not verified by OpenAI.

External References

Contribution & Novelties

The video offers a practical, accessible guide to deploying open-source AI models locally, addressing data privacy concerns and cost savings. It bridges the gap between news about DeepSeek and actionable steps for users to take control of their AI tools.

Pour aller plus loin :

  • Ollama — Essential tool for running open-source models locally.
  • Open WebUI — User-friendly interface for interacting with local models.
  • DeepSeek R1 paper — Academic paper on DeepSeek R1, providing technical details.
  • Hugging Face — Platform for discovering and sharing open-source models.
  • n8n — Open-source automation tool, mentioned in the video.

95 words

Radar Profile

The radar profile shows moderate scores across all dimensions, with a slight emphasis on information quantity and technical level, reflecting the video's practical tutorial nature. The lower reliability score indicates the lack of rigorous sourcing and promotional content.

Reliability 5/10

💬 Très positif. Sur les 30 commentaires analysés, les spectateurs expriment un fort enthousiasme pour le contenu et demandent davantage de tutoriels sur les outils open source, sans critiques négatives.