[NFBCS] Local AI

Thu May 21 18:10:20 UTC 2026

Hi Paul, and all,

Thank you for the comprehensive response and subsequent discussion.

My goal is twofold: First, try to save a little money now that the usage is
moving more toward a metered standard. Right now I’m using Perplexity Pro
for academic research and Gemini Pro for everyday queries. I think the time
saved has made both subscriptions a worthwhile investment. I just recently
began with Claude and like that I’ve been able to train it to write in my
style and tone. Perhaps Gemini and Perplexity allow for similar capability,
but Claude has made this a bit more intuitive and has resulted in excellent
working drafts I can use to finalize what I’m going for. I think I’ll
likely end up biting the bullet and paying for a subscription there as
well, but the type of subscription is what leads to my second goal.

My second goal is to move more into the agentic space. There are a set of
repetitive tasks I would just assume dump onto a virtual assistant. I was,
still am, on the fence about learning Python to tackle these tasks, and if
I hesitate at all, it is because I fear diving into an area that does not
really contribute to my optimal productivity. Coding would be great in
terms of understanding the underlying method that created what I needed,
but in the end, a coder is not what I need to be a part of my core set of
skills.

But, the idea of maintaining spreadsheets based on PDF files received and
using both to create reports and invoices is highly appealing. This is
working with data I would rather not share on an open web.

After all this, there are marketing, creative projects, and other
proprietary data that I would like to keep offline where possible. It’s not
data so sensitive that I will absolutely refuse to put it through the
online systems, but if I could keep it offline and rely on a reasonably
fast response, that would be excellent.

Based on what I’m reading here, it sounds like my offline dream is not
going to come true. My workstation may not be up for running an LLM strong
enough to engage with me in the way I need. If I’m reading that
incorrectly, I’m happy to be told otherwise.

To summarize, at this stage, I am not doing any coding. Although I am
tempted by the vibe coding trend, I understand it is always better to
understand the programming going on so I can fix the errors myself. I am
mostly relying on document analysis, research, and coaching to help turn my
outlines into usable drafts. I do some image generation and occasionally
ask Gemini to review audio/video files, but I’m okay with keeping those
firmly in the paid subscription buckets.

I hope this has made some kind of sense and hope that my questions are
helping other non-technical people figure out how to best maximize the
available options. The open weight options sound really good on paper, but
I don’t want to go through the trouble of installing local models if my
machine won’t make a substantial difference and if, in the end, I would be
better off making the best of the online models.

Very appreciative,

Joe

On Thu, May 21, 2026 at 1:36 AM Paul York via NFBCS <nfbcs at nfbnet.org>
wrote:

> I've been knee deep in local llm setup for the better part of the last two
> weeks. To give you perspective on my hardware, I'm playing with two
> moderately beefy "consumer" machines: a Windows 11-based Ultra 7, 64GB RAM,
> RTX 4070 w/ 12GB VRAM and a linux-based Ryzen AI 9 HX370 mini pc with 64GB
> RAM (both bought before prices went bonkers thankfully).
>
> TLDR: I'm keeping my Claude and Gemini subscriptions.
>
> I think a longer discussion will hinge on what you want to do with it. Are
> you programming? Running OpenClaw/agentic stuff? Just chatting? Creating
> documents and presentations? Doing NotebookLM kind of things? Because
> here's the deal. After a LOT of tweaking, I'm getting:
>
>    - around 25 tokens per second output on my iGPU (Ryzen) using some
>    pretty high quality models (Qwen 3.6 35b and Gemma 4 26b) by pushing the
>    VRAM up to 48GB.
>    - anywhere between 65 and 95 tokens per second output on my RTX GPU
>    using much lower quality models (Qwen 3.5 9b and Gemma 4 4b).
>
> In both cases, if I don't take steps to optimize the model such that it
> stays 100% in VRAM, it slows to an entirely unusable rate.
>
> UP FRONT WARNING--I'm a noob with this, so take my explanation with a
> grain of salt.
>
> What does that actually mean? Well especially if you use a "reasoning"
> model like Qwen, then a simple query response (like "tell me a funny dad
> joke") can take up to a minute to respond. This is because approximately
> every word of every "thought" is an output token. It "talks to itself"
> until if decides it has found a reasonable answer. And it adds up quick.
> Here are some basic results for this exact query on all 4 models / hardware:
>
>    - Qwen on RTX 4070: required 1200 tokens and 18 seconds to respond
>    - Gemma on RTX 4070: required 300 tokens and 3.5 seconds to respond
>    - Qwen on iGPU: required 630 tokens and 23 seconds to respond
>    - Gemma on iGPU: required 450 tokens and took 18 seconds to respond
>
> Again that's moderately beefy hardware and a lot of tweaking. But I could
> also do far better if I accepted much dumber models. Which may be just
> find for basic agentic work. But much less good for coding or reasoned
> synthesis. And the "smartest" model took 23 seconds to reason through a dad
> joke. 7 seconds just to figure out how to respond to "hello". Working on
> truly complex reasoning can take a bathroom+coffee break to give you back
> results.
>
> Note too that this doesn't take into account context size and context
> caching. Context is the LLM's active memory. Models have maximums (I think
> they are tuned for these sizes). But in most cases you'll likely have to
> accept something lower. However, to be even moderately useful for much of
> anything, you can't go terribly low. Coding tools and agentic tools just
> blow up if they can't remember things from one thought to the next.
>
> The numbers I'm getting on the RTX are decent. Almost usable. BUT the
> context sizes to achieve that make it basically unusable for the kind of
> work I want to do. If I bump up the context window to a usable level with
> these models, I leak out into RAM (far slower than VRAM) and my performance
> tanks to unusable levels (like < 5-10 tps...at least 80-90% or more slower).
>
> The iGPU with huge VRAM is slower than the dedicated GPU, but because I
> can crank up the context window, they actually become usable for what I
> want to use them for. However...speed. Claude Sonnet or Gemini Flash are
> easily 10x faster at everything. And more like 20x-30x faster for most
> reasoning work. So at some point it's a question of how much you value your
> time.
>
> I will be using local models for some basic stuff, I think. I'm starting
> down a personal knowledge management path with AnythingLLM or something
> similar. I think it'll pair perfectly with this. And I'll likely find more
> ways to leverage it. But I won't be abandoning the big boys any time soon.
>
> And sadly, although your Ultra 7 w/ 32GB RAM is an awesome PC, I fear your
> experience with local llms for anything other than experimentation and
> learning will prove frustratingly slow. And with prices the way they are
> right now, getting your PC spec'd to perform moderately well will certainly
> cost around the same as a full year of one of the "ultimate" plans.
>
> Hope this was helpful. And that I didn't show my ignorance too badly.
>
> Best,
> Paul York
>
> On Wed, May 20, 2026 at 11:35 PM Lewis Wood via NFBCS <nfbcs at nfbnet.org>
> wrote:
>
>> I am currently learning as well.
>>
>>
>>
>> I am now doing Ollama playlist lessons #2 currently.
>>
>> https://www.youtube.com/playlist?list=PLvsHpqLkpw0fIT-WbjY-xBRxTftjwiTLB
>>
>>
>>
>>
>>
>> I did my initial research on Lm Studio before I learned about Ollama CLI
>>
>>
>>
>> This was my first Lm Studio and it was an excellent one regarding
>> resources, models, agents, etc.  Even discussed how to load partial in
>> differing areas gpu and ddr.
>>
>> https://www.youtube.com/watch?v=UngVdAsQEiU
>>
>>
>>
>>
>>
>> You can search youtube  “lm studio”
>>
>>
>>
>> Lewis Wood
>>
>>
>>
>>
>>
>>
>>
>> *From:* NFBCS <nfbcs-bounces at nfbnet.org> *On Behalf Of *Joe Orozco via
>> NFBCS
>> *Sent:* Wednesday, May 20, 2026 10:08 PM
>> *To:* 'NFB in Computer Science Mailing List' <nfbcs at nfbnet.org>
>> *Cc:* Joe Orozco <jsorozco at gmail.com>
>> *Subject:* [NFBCS] Local AI
>>
>>
>>
>> Hello,
>>
>>
>>
>> With Google following in Claude’s footsteps in terms of usage
>> restrictions, can anyone speak to their experience using local LLM options?
>> I’m looking at Jemma 4 and trying to understand how accessible this route
>> might be with JAWS on Windows.
>>
>>
>>
>> I’m on a fairly decent machine: 32 GB RAM, Ultra 7 processor, 4 TB SSD. I
>> see they’re recommending GPU for some of the more robust models, but I want
>> to think most of what I’m doing shouldn’t require gaming machine specs. If
>> you beg to differ though, let me know.
>>
>>
>>
>> If anyone can speak to Jemma alternatives, I’d also be interested. I
>> don’t think I’ll suspend my subscriptions, but with these usage limitations
>> feeling like the new standard, I want to spread my usage a little so that I
>> don’t feel like I need to be hitting the top subscriptions just to get more
>> mileage out of the five-hour increments.
>>
>>
>>
>> Thanks in advance for any tips,
>>
>>
>>
>> Joe
>>
>>
>>
>> --
>>
>> Joe Orozco: Your Message, My Mission
>>
>> https://joeorozco.com/services/
>>
>>
>> _______________________________________________
>> NFBCS mailing list
>> NFBCS at nfbnet.org
>> http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
>> To unsubscribe, change your list options or get your account info for
>> NFBCS:
>> http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/paul%40yorkfamily.com
>>
> _______________________________________________
> NFBCS mailing list
> NFBCS at nfbnet.org
> http://nfbnet.org/mailman/listinfo/nfbcs_nfbnet.org
> To unsubscribe, change your list options or get your account info for
> NFBCS:
> http://nfbnet.org/mailman/options/nfbcs_nfbnet.org/jsorozco%40gmail.com
>

-- 
--

Joe Orozco: Your Message, My Mission
https://joeorozco.com/services/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nfbnet.org/pipermail/nfbcs_nfbnet.org/attachments/20260521/ff189a5f/attachment.htm>