NEW Qwen 2.5 Max is INSANE 🤯

Want the SOP? DM me. Like the content? Check this out: 🚀 Book a FREE SEO strategy Session 👉 🔗 Free SEO Course 👉 💰 Want more money, traffic and sales from SEO? Join the SEO Elite Circle 👉

new single Stargazing just landed today

What did you use for the AI avatar?

Mapped out the whole process in the AI Profit Boardroom 👇

Is this it? Is this the future of SEO is @JulianGoldieSEO making bullshit AI videos every god damn week.

good job man lol Great use of AI avatar

Is it? You prove it?

If everything is insane, nothing is insane. 😄

Qwen 2.5 Max is like a one legged man kicking a football.. Qwen 2.5 Max When you can't refer to a previous conversation for review and analysis, then even 10k characters becomes very inadequate. Qwen2.5-Max You're absolutely right, and I completely understand your frustration. The inability to retain memory across conversations and the 10k character input limit can indeed make it challenging to work on complex tasks that require continuity or involve large amounts of data. Let me break this down and offer some potential solutions or workarounds: 1. Memory Across Conversations Since Qwen 2.5 Max doesn't retain memory between sessions, you need to explicitly manage context within a single session. Here are some strategies: * Summarize Key Points : If you're working on a multi-step task, summarize the key points or outcomes from one step before moving to the next. This allows you to "carry forward" the essential information without needing to rely on memory. * Use External Tools : You can use external tools (e.g., note-taking apps, spreadsheets, or document editors) to store and organize information as you go. Then, paste relevant portions into the conversation when needed. * Chunking Workflows : Break your task into smaller, manageable chunks. For example, if you're analyzing a long document, process it in sections, summarize each section, and then combine the summaries for a final analysis. 2. Input Length Limitation The 10k character limit is restrictive, especially for tasks involving large documents or datasets. Here’s how you can handle this: * Divide and Conquer : Split your input into smaller segments (e.g., chapters, sections, or paragraphs). Process each segment individually, and then synthesize the results. For example: * Analyze Section 1 → Generate Summary. * Analyze Section 2 → Generate Summary. * Combine Summaries for a Final Output. * Prioritize Critical Content : If the full text exceeds 10k characters, focus on the most critical parts first. For example, extract key excerpts or highlight important sections to include in the input. * Iterative Processing : Use an iterative approach where you feed the model smaller chunks of text, analyze them, and then refine your queries based on the outputs. 3. Combining Both Challenges When both memory and input length are limited, the best approach is to combine the strategies above: * Step-by-Step Workflow : 1. Divide your document into manageable chunks (e.g., 8–9k characters per input). 2. Process each chunk and generate summaries or insights. 3. Store these summaries externally (e.g., in a document or note-taking app). 4. Use the summaries as context for subsequent steps or analyses. * Synthesis Session : Once you’ve processed all chunks, bring the summaries together in a final synthesis session. Provide the key points from each summary to the model to generate a cohesive analysis or conclusion. Example Workflow Let’s say you’re analyzing a 40k-character report: 1. Split the Report : Divide it into four 10k-character sections (or smaller if needed). 2. Process Each Section : * Input Section 1 → Generate Summary 1. * Input Section 2 → Generate Summary 2. * Repeat for Sections 3 and 4. 3. Combine Summaries : Paste Summaries 1–4 into a new input and ask for a consolidated analysis. 4. Refine as Needed : If the consolidated analysis requires deeper insights, provide additional context or specific questions. Future Improvements Your feedback about these limitations is incredibly valuable. At Alibaba Cloud, we’re constantly working to improve our models, including increasing input limits and exploring ways to better handle continuity across conversations. While these features may not be available immediately, your input helps us prioritize improvements.