Here’s o3 mini on an internal benchmark:
GENERATIVE AI
-
o3-mini vs Sonnet: Speed tradeoffs in AI models
By
–
We’ve had o3-mini internally @every for a few days. Here’s the vibe check from Cora GM @kieranklaassen
: pros: It’s like o1 but faster
cons: Sonnet still better for html, css, and UX -
OpenAI o3-mini: Faster, cheaper model for coding and agentic apps
By
–
Welcome, @OpenAI o3-mini! A fantastic model for coding and building agentic apps, complete with function calling, Structured Outputs, and streaming. It outperforms o1 on medium/high reasoning tasks, is much faster, and costs 93% less. Can’t wait to see what you build! https://
x.com/OpenAIDevs/sta
/OpenAIDevs/status/1885407761644609776
… -
O3-mini outperforms o1 in software engineering tasks
By
–
o3-mini (high) beating o1 at software engineering is a win for everyone. The trend is better, faster, cheaper. Excited to see how o3 pro compares.
-
OpenAI Partners with National Labs for AI Science Research
By
–
Anthropic CEO says AI model costs remain high across companies, while national labs gain AI access for science research, and Google brings AI analytics to everyday spreadsheet users. Here's your daily AI news briefing for January 31st, 2025: OpenAI Partners with National Labs
-
DeepSeek on SambaNova Cloud Accelerates AI Agent Development
By
–
Check out @deepseek_ai on SambaNova Cloud! Accelerate your AI & ML workflows, unparalleled speed & efficiency Use our speed with #DeepSeek to build the fastest #AI Agents Integrations with @HuggingFace & @continuedev to code faster Transform your work
-
AI’s Trajectory: From Knowledge Work to Superintelligence
By
–
And even crazier: If we extrapolate from AI’s current trajectory, its impact on knowledge work is obvious. But that’s just the beginning. This leads to another revolutionary phase: once we’ve built an 'AI brain' that’s smarter than us—one that can reason, 'think', and
-
Deepseek-R1 Now Available on Mosaic AI Model Serving
By
–
We're excited to share that you can now run Deepseek-R1 on Mosaic AI Model Serving. Deepseek-R1, with its reasoning capabilities, has changed the narrative around open-source #LLMs and is a big milestone in democratizing AI. With the Databricks Data Intelligence Platform, our
-
Company Shares AI Training Knowledge Openly This Week
By
–
since openly sharing AI-training knowledge is in the air this week, we decided to go all in
-
China integrates 81 LLM contracts in military and government sectors
By
–
2/ in the first half of 2024, the Chinese government issued 81 contracts to LLM companies to integrate it into their military and government, with Zhipu emerging as the early winner