9 months ago
DeepSeek Launches New AI Models to Rival Gemini and ChatGPT
DeepSeek, a Chinese company, has launched two new AI models called DeepSeek-V3.2 and DeepSeek-V3.2-Speciale.
These models are designed to compete with popular AI models like Gemini and ChatGPT.
DeepSeek claims that their new models perform as well as GPT-5 in general tasks and even better than GPT-5 High and Gemini 3 Pro in certain areas like math and coding.
The models have also achieved top scores in important math competitions.
To try these new models, you can simply log in to the DeepSeek website or app.
DeepSeek launches DeepSeek-V3.2 and DeepSeek-V3.2-Speciale to compete with Gemini and ChatGPT.
DeepSeek-V3.2 claims GPT-5 level performance in general tasks.
DeepSeek-V3.2-Speciale matches Gemini 3 Pro in complex problem-solving.
DeepSeek-V3.2-Speciale achieves Gold-medal in 2025 IMO and IOI.
DeepSeek-V3.2-Speciale outperforms GPT-5 High and Gemini 3 Pro in math and coding benchmarks.
- Who
- DeepSeek, a Chinese AI startup
- What
- Launched new AI models, DeepSeek-V3.2 and DeepSeek-V3.2-Speciale
- Where
- China
- When
- Not specified
- Why
- To compete with Gemini and ChatGPT in the AI chatbot space
Key facts
- Company
- DeepSeek
- New Models
- DeepSeek-V3.2 and DeepSeek-V3.2-Speciale
- Performance Claim
- GPT-5 level in general tasks
- Specialization
- Reasoning-first models built for agents
- Benchmarks
- Gold-medal in 2025 IMO and IOI, outperforms GPT-5 High and Gemini 3 Pro in math and coding
- Availability
- Default model on DeepSeek's website and app
Quotes
DeepSeek
A Chinese AI startup based in Hangzhou.
“DeepSeek-V3.2 is our first model to integrate thinking directly into tool-use, and also supports tool-use in both thinking and non-thinking modes.”
thehindubusinessline.com
“DeepSeek-V3.2 achieves similar performance with Kimi-k2-thinking and GPT-5 across multiple reasoning benchmarks.”
thehindubusinessline.com
Sources
Related news
Alphabet Shares Slide as AI Talent Departures Fuel Investor Concerns
Human-Like AI Chatbots May Influence Personal Identity, Study Finds
AI Chatbots Improve Suicide-Risk Responses, Yet Safety Gaps Remain
Google’s Pixel 11 Pro Fold Makes Foldables Everyday Phones
Google Pixel 11 Reviews Put Software and AI Ahead
