Back

Rahul Agarwal

Founder | Agentic AI... • 14h

Anthropic just released Claude Opus 4.6. Here’s what’s new: 1) Smarter problem solving. It tackles complex tasks efficiently and doesn’t waste compute on simple ones. 2) 1M token context window. Enough to hold roughly 10 full novels in one session. 3) Long context works effectively. Opus 4.6 scores 76% on retrieval tasks compared to 18.5% for Sonnet 4.5. 4) Outperforms GPT 5.2 on real world tasks like finance, legal, and research with a 144 Elo point lead. 5) Can debug its own code, reducing the need for manual guidance. 6) Agent teams let multiple AI agents work simultaneously on different code sections. 7) Four effort modes. Choose the balance between speed and quality: low, medium, high, max. 8) Leading model for uncovering hard to find online information. 9) Direct integration in Excel and PowerPoint. Analyze data and build presentations without switching apps. 10) Pricing remains $5/$25 per million tokens. You get a more capable model at no extra cost. Share it with others so even they can also know about this update. Follow me for more such information on AI.

Reply
1

More like this

Recommendations from Medial

Image Description

Yash K

Avid Learner • 17h

When I have children, I will tell them a story. On Feb 5, 2026, Opus 4.6 and Codex 5.3, two of the best models, launched within 30 minutes of each other. It was pure chaos. Opus 4.6 was expected, but Codex 5.3 was a surprise and it absolutely nailed

See More
Reply
2
1

Kimiko

Startups | AI | info... • 8m

✨ Claude Opus 4 = World's best coding model (72.5% on SWE-bench) ✨ Claude Sonnet 4 = Major upgrade, powering GitHub Copilot's new agent ✨ Both can think + use tools simultaneously ✨ Claude Code now GA with VS Code/JetBrains integration

Reply
9

Soham Khadse

 • 

PTC • 1m

If you want to have Opus 4.5 at it's MAX in Claude Code Update ~/.claude/settings.json { "env": { "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "64000", "MAX_THINKING_TOKENS": "31999" } } - double the output - ultrathink always on Ctrl+O will show

See More
Reply
1

Rishabh Raj Pathak

Polymath • 8m

OpenAI’s new Codex is here! Now in ChatGPT Pro/Enterprise/Team (preview) + CLI, it automates coding tasks-writing features, fixing bugs, running tests-in a secure sandbox. - Automates everyday coding: writes new features, fixes bugs, and generate

See More
Reply
3

Baqer Ali

AI agent developer |... • 4m

🚨 BREAKING: Anthropic just dropped the tutorial on "Building AI Agents with Claude Agent SDK" Here's what it covers: > Agent Loop Gather context → Take action → Verify work → Repeat. Your agent searches files, executes tasks, checks its output, t

See More
Reply
1
12

Comet

#freelancer • 1y

Using AI tools can significantly enhance productivity by 10 times. Here are some examples of AI tools that can help with different tasks: - Instead of manually coding, you can use a code interpreter, which is an AI tool that can interpret and execut

See More
Reply
3
Image Description

Yogesh Jamdade

..... • 10m

how to connect ai models with your own softwares or tools? The answer is mcp servers The Model Context Protocol (MCP) is an open standard that enables AI models to securely access data and tools through standardized servers. MCP servers facilitate

See More
1 Reply
2
Image Description
Image Description

Vishu Bheda

 • 

Medial • 7m

𝗔𝗜 𝗶𝘀𝗻’𝘁 𝗮𝗯𝗼𝘂𝘁 𝘄𝗵𝗶𝗰𝗵 𝘁𝗼𝗼𝗹 𝘆𝗼𝘂 𝘂𝘀𝗲. It’s about how you use the right ones together. 📌 𝗖𝗵𝗮𝘁𝗚𝗣𝗧: for deep thinking 📌 𝗖𝗹𝗮𝘂𝗱𝗲: for clean, reliable code 📌 𝗣𝗲𝗿𝗽𝗹𝗲𝘅𝗶𝘁𝘆: for best research 📌 𝗡𝗼𝘁𝗲𝗯𝗼𝗼�

See More
4 Replies
3
13

Nandeesh N

For the greater good... • 6m

HOW TO VIBE CODE EFFECTIVELY Step 3: Act Like a Product Manager for the AI Your primary role when vibe coding is to be the product manager for your AI assistant, This means: Provide Clear Guidance and Context: Just as you would with a new human emp

See More
Reply
8

Download the medial app to read full posts, comements and news.