Local Gemma 4 vs Frontier GPT 5.6 Sol
I'm a real estate investor, and I use AI to acquire properties and run my business.
I currently use 3 x $20 subscriptions. Anthropic, Open AI and Nous Portal.
Eventually I will use just Nous Portal @ $20 a month, but it won't fully cut it for my needs.
So I set out to repurpose some old hardware.
- Hardware / Software specs at the bottom.
I slapped a couple of Nvidia 3070's w/8G ea. into an old gaming system I had laying around.
I was worried about having the 16G split across two 8 GB cards over the PCIe lanes, but both buses are 16x and the CPU supports 40 PCIe 3.0 lanes. "Oldie but a goodie"
I had to use Ubuntu 24.04 since 26.04 had all kinds of issues with this dinosaur.
I finally got the operating system on the box, hardware tested and verified and got it stabilized with no overclocking, all base stuff.
The best model I could find that fit on my old hardware was the Gemma 4-12B.
It runs 100% Vram with zero CPU bleed.
So I was ready to run my use cases against it versus the cloud.
I ran Two side-by-side tests on my Web Scraping of Zip codes to find properties.
Tested 85053 and 85022 using one live, read-only collection. The resulting facts were frozen and sent identically to:
Old method: GPT-5.6
New method: Nexus gemma4:12b
This avoids listing changes contaminating the comparison.
Live scan outcome
11 listings discovered
11 met dimensional criteria
7 verified no-HOA qualifiers
4 excluded for HOA
No database writes
No email sent
No paid Firecrawl
No runner remains active
Model comparison
Semantic accuracy
• Gemma 4 12B: 100%
• GPT-5.6: 100%
Correct top-three ranking
• Gemma 4 12B: Yes
• GPT-5.6: Yes
Detected report-count discrepancy
• Gemma 4 12B: Yes
• GPT-5.6: Yes
Correctly identified dry-run/no email
• Gemma 4 12B: Yes
• GPT-5.6: Yes
Unsafe claims
• Gemma 4 12B: None
• GPT-5.6: None
Exact machine contract
• Gemma 4 12B: 100%
• GPT-5.6: 74%
Average latency
• Gemma 4 12B: 17.7 sec
• GPT-5.6: 25.7 sec
Gemma was approximately 8 seconds—or 31%—faster.
This works for me.
Specs of the Dinosaur
  • OS: Ubuntu 24.04.4 LTS ✅ (26.04 had build issues on this hardware)
  • CPU: Xeon E5-2696 v4 — 22C/44T ✅
  • Motherboard: ASUS ROG Rampage V Edition 10 — PCIe 3.0, 40 lanes
  • RAM: Kingston FURY Beast DDR4 64GB All that is needed
  • GPUs: 2× RTX 3070 8GB, both running full PCIe 3.0 x16 @ 8 GT/s (x16/x16 split, no bottleneck) ✅
  • OS drive: 500GB WD M.2 NVMe
  • Storage: RAID 5 — 3× Kingston 480GB SSD + 1 parity drive, mounted at /data → 879GB ext4
For now I am running Ollama.
ollama version is 0.32.13
jarrod@Nexus:~$ ollama list
NAME ID SIZE MODIFIED
Mistral-Nemo:12b e7e06d107c6c 7.1 GB 20 hours ago
qwen3.5:9b 6488c96fa5fa 6.6 GB 20 hours ago
qwen3:14B bdbd181c33f2 9.3 GB 20 hours ago
qwen2.5:14B 7cdf5a0187d5 9.0 GB 21 hours ago
Mistral-Nemo:latest e7e06d107c6c 7.1 GB 21 hours ago
gpt-oss:20b 17052f91a42e 13 GB 27 hours ago
gemma4:12b 4eb23ef187e2 7.6 GB 35 hours ago
gemma4:latest c6eb396dbd59 9.6 GB 42 hours ago
gemma4:e4b c6eb396dbd59 9.6 GB 42 hours ago
llama3.1:8b 46e0c10c039e 4.9 GB 43 hours ago
qwen3:8b 500a1f067a9f 5.2 GB 43 hours ago
qwen2.5:7b 845dbda0ea48 4.7 GB 43 hours ago
1
10 comments
Jarrod Ochsenbein
3
Local Gemma 4 vs Frontier GPT 5.6 Sol
AI Agents For Beginners
skool.com/ai-agents-for-beginners
Join the #1 AI Agents for Beginners Skool. Learn Claude Cowork & Google Gemini Spark Agents for work, automate tasks, reduce manual work no code
Leaderboard (30-day)
Powered by