Ian Piggott

hmmm I knew it seemed off. Q. compared to grok 4.5, does 4.6 use more tokens/allowance? seems I'm burning through my usage quicker A. Yes, your observation is correct: Grok 4.6 burns through your usage allowance much faster than Grok 4.5. While the baseline sticker price remains exactly the same ($2 per million input tokens and $6 per million output tokens), Grok 4.6 uses roughly 30% more tokens per task on average due to architectural changes and advanced internal processing. Think I'll investigate the 'sweet spot' area. see below...

New comment 1d ago

Ian Piggott

0 likes • 1d

I had to use Google Gemini to get that answer, as Grok refused to answer it. Grok was unable to finish replying. Please try again later or use a different model. There is further expansion on this, if one changes parameters, like High vs Med, 4.5 vs 4.6, etc.

Ian Piggott

0 likes • 1d

I think I maybe just be pushing it harder.... 🤣

Shane McGrath

26d •

General discussion

How many of you guys are on X? Drop your handles in here. I’m @vibecodeguild

New comment 20h ago

Ian Piggott

2 likes • 1d

https://x.com/ian_piggott

Ikbel Hammami

7d •

General discussion

Feedback on the current AI models

Does anyone have some feedback on the Grook Vs Claude Vs Perplexity ?

New comment 2d ago

Ian Piggott

1 like • 2d

EDITS: grammar / spelling. Grok Build in TUI (terminal user interface) has been and is my only deep foray into AI agents (paid plans - SuperGrok Heavy) I have just been building websites (around 80), and business planning, for the last 6 weeks or so. It has been great for this. I adjust the model so I don't waste tokens/allowance on trivial/mundane tasks that a lower AI model can do. Business planning and scope/roll-outs have been on med to high models of Grok (Grok 4.5 high then Grok 4.6 med-high. I rarely waste my allowance on Grok 4.6 xhigh) I want to do much more with it, but I have a task that is very large and all my spare time is deidcated to that until completion. Then I'll try some apps building and other projects I have planned. The only other AI's I use would be the free ones like claude (sonnet 5 extra), Google Gemini, (which actually does some good work), Brave Leo (not sure the LLM it uses), and ChatGPT (which did a excellent / solid business plan breakdown on it free plan. I was super impressed) That's where I'm at so far. I like Grok for what it is right now and have no good or financial reason to change. I'm looking forward to 4.7 and 5.0 when they are release. 5.0 is to be a "banger" from the limited reports and info of it.