cd /news/large-language-models/turintech-s-artemis-shrinks-google-s… · home › topics › large-language-models › article
[ARTICLE · art-171907] src=alphasignal.ai ↗ pub= topic=large-language-models verified=true sentiment=↑ positive

TurinTech's Artemis Shrinks Google's Gemma 4 E2B to 2.05 GB Without Custom Code

TurinTech released three Artemis GGUF builds of Google's Gemma 4 E2B chat model that compress the 9.3 GB reference checkpoint to between 2.05 GB and 2.31 GB, a 75% to 78% reduction, while the default Q3_K_L file matches the original model's next token 95.8% of the time. TurinTech's benchmarks, run with stock llama.cpp on an AMD Ryzen 9 9950X CPU, show other public quantizations below 2.6 GB lose 23 to 34 task points, whereas the Artemis builds range from −1.6 to +0.3 points on a 310-item evaluation set. The builds are free under Apache-2.0, text only, and run on stock llama.cpp, Ollama and LM Studio, though TurinTech reports longer outputs drift and small math losses, and Unsloth's 2.62 GB UD-Q4_K_XL file remains more faithful at 98.3% next-token agreement.

key takeaway

TurinTech's Artemis Q3_K_L build cuts Gemma 4 E2B from 9.3 GB to 2.31 GB while matching the reference model's next token 95.8% of the time and losing 1.6 task points, versus 23 to 34 points lost by other public quants below 2.6 GB.

read2 min views3 publishedNov 18, 2026
TurinTech's Artemis Shrinks Google's Gemma 4 E2B to 2.05 GB Without Custom Code
Image: Alphasignal (auto-discovered)

TurinTech's Artemis GGUF build packs Google's Gemma 4 E2B into 2.05 to 2.31 GB, and it holds up much better than other public files in that size range.

  • TurinTech released Gemma 4 E2B Artemis GGUF : three files, 2.05 to 2.31 GB, down from 9.3 GB.
  • The default Q3_K_L file matches the original model's next token 95.8% of the time.
  • Other public quants of the same checkpoint below 2.6 GB lose 23 to 34 task points.
  • The trick: keep Google's QAT Q4_0 weights and compress only the per-layer embedding tables harder.
  • Free under Apache-2.0, text only, runs on stock llama.cpp, Ollama and LM Studio.
  • Weak spots: longer outputs drift, small math losses, no memory measurements yet; Unsloth's 2.62 GB file is more faithful.

TurinTech, the company behind the Artemis code optimisation platform, has released three compressed builds of Google’s smallest Gemma 4 chat model in its Artemis GGUF repository. The 2.05 GB to 2.31 GB files are 75% to 78% smaller than the 9.3 GB reference checkpoint.

All three builds use the standard GGUF format and run with stock llama.cpp, without forks or custom kernels. TurinTech’s benchmarks show them retaining more of the original model’s behaviour than the other public quantizations it tested at similar sizes.

The public baseline drops sharply below 2.6 GB #

At 9.3 GB, the reference checkpoint exceeds the physical memory of an 8 GB phone before accounting for the operating system, application or inference overhead. Quantization reduces that footprint by storing model weights with fewer bits, although aggressive compression can change the model’s token choices and reduce task accuracy.

TurinTech benchmarked each file against the reference checkpoint using the same llama.cpp settings on an AMD Ryzen 9 9950X CPU. Unsloth’s 2.62 GB UD-Q4_K_XL remained the closest public build. Other tested builds below that size chose a different next token 45% to 55% of the time and lost 23 to 34 percentage points across the sampled tasks.

The reported next-token metric checks whether a build selects the same highest-probability token as the reference. The 32-token metric compares their generated prefixes, while task accuracy measures outcomes across a 310-item evaluation set.

TurinTech-reported results against the full-size reference checkpoint
Build Size Next-token agreement 32-token match rate Task accuracy change
--- --- --- --- ---
Unsloth UD-Q4_K_XL 2.62 GB 98.3% 89.1% −0.3 points
Artemis Q3_K_L, default 2.31 GB 95.8% 74.7% −1.6 points
Artemis Q3_K_M 2.20 GB 93.1% 69.0% +0.3 points
Artemis IQ3_XXS 2.05 GB 91.6% 67.1% 0.0 points

This story is for Pro members #

You've reached the end of the free preview. Upgrade to AlphaSignal Pro to read the full article - and everything else behind the paywall.

── more in #large-language-models 4 stories · sorted by recency
── more on @turintech 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/turintech-s-artemis-…] indexed:0 read:2min 2026-11-18 · —