O3
o3
O3 is an AI reasoning model developed and released by OpenAI, positioned as the company's most advanced reasoning system to date . OpenAI introduced it in December 2024 alongside the smaller o4-mini, with both models becoming widely available in April 2025 . The company described o3 and o4-mini as likely the final standalone reasoning models before the anticipated arrival of GPT-5, which is expected to unify traditional and reasoning architectures .
In benchmark terms, o3 made notable waves by achieving an 87.5% score on the ARC-AGI challenge—a test that had sat unsolved for five years—using a high-compute configuration, and it posted state-of-the-art results on Codeforces, SWE-bench, and MMMU . Per 云启资本's review, it also reduced major errors by roughly 20% compared with o1 in difficult real-world tasks spanning programming, consulting, and creative work . A distinctive feature of o3 and o4-mini is their "image reasoning" capability: they can manipulate images directly within their chain of thought—zooming, rotating, and otherwise adjusting visuals as part of the reasoning process—and they can autonomously invoke ChatGPT tools such as web browsing, Python execution, and image generation .
On pricing and access, o3 and o4-mini undercut their predecessors at equivalent tiers, with o3-mini notably 63% cheaper than o1-mini; both models offer a 200,000-token context window in the API . 真格基金's Yusen Dai, writing in mid-2025, judged o3 to represent a "clear step up" from o1 in practical use, with o3 pro showing further gains in reasoning length and logical coherence, even if the underlying Transformer architecture remained an evolution rather than a paradigm shift .
AI-generated — may contain errors, please verify.
Coverage
Are Models and "Shells" Both Undervalued? ZhenFund's Yusen Dai's Mid-2025 AI Review --- The past six months have been a turbulent period for AI. On one hand, DeepSeek's breakthrough has reignited global enthusiasm for Chinese AI; on the other, the commercialization path for large models remains unclear, and the market is growing increasingly anxious about when AI will generate real returns. Against this backdrop, I believe it's necessary to take stock of where we stand at this mid-point in 2025. **The Core Question: Are Models and "Shells" Both Undervalued?** Over the past two years, the AI industry has cycled through several narratives: from the initial frenzy over foundation models, to the subsequent boom in AI applications ("shells"), then a period of disillusionment as applications failed to deliver, and now a renewed focus on models following DeepSeek's success. But I want to propose a contrarian view: **both models and "shells" may be simultaneously undervalued by the market right now.** This might sound paradoxical. Let me explain. **Why Models Are Undervalued** The market's current skepticism toward model companies stems largely from two concerns: the commoditization of models (will they all become undifferent
In 2025, the "Sedol moment" for industries far and wide is only just beginning.
Quick Take on OpenAI's New Models: o3, o4-mini — What Can Image Reasoning + Autonomous Tools Bring? | Yunqi Tech π
In the early hours of April 17 Beijing time, OpenAI released two new AI reasoning models, o3 and o4-mini. These are likely the last standalone AI reasoning models before GPT-5, with further improvements in both performance and cost efficiency. In this episode of **"Yunqi Tech π"**, we break down the details for you.
A Deep Dive into the Beauty of DeepSeek's Creation: How Was DeepSeek R1 Forged?
AI penetration today is only 5%. For the remaining 95%, what will their first AI application be?
2025 New Year Conversation with Koji: AI's Pivotal Year, the Dawn of the Agent Era | Seriously Speaking EP35
Stay optimistic about AI.
A New Year Conversation with Koji: 2025 — The Critical Year for AI, the Dawn of the Agent Era
2025 Kickoff Conversation: AI's Pivotal Year, the Dawn of the Agent Era | A Conversation with Yusen Dai of ZhenFund
The Evolution of Coding Agents: In-Depth Conversations with Chinese and American Agent Founders, Alibaba Researchers, and Investors





