🇺🇸 Gemma 4 12B
Gemma 4 12B (Google): dense multimodal model (text, vision, audio), 256k context, ~7 GB Q4 VRAM. Apache 2.0, multilingual.
# HuggingFace: google/gemma-4-12B
Rankings updated on 22/10/2026
The RTX 4070 Super (12 GB GDDR6X, 504 GB/s) boosts the 4070 by ~15% in CUDA cores. Excellent performance-per-price ratio for 7–14B models in Q4_K_M.
RTX 4070 Super: purchasing alternative available for local AI — RTX 5060 Ti 16 GB:
Which PC should you choose for your budget? Our picks from €800 to €3,500 →
Affiliate links — BestLLMfor may earn a commission on purchases, at no extra cost to you, which does not influence the ranking (established independently). As an Amazon Associate, BestLLMfor earns from qualifying purchases.
Gemma 4 12B (Google): dense multimodal model (text, vision, audio), 256k context, ~7 GB Q4 VRAM. Apache 2.0, multilingual.
# HuggingFace: google/gemma-4-12B
GLM 5.3 (Zhipu): dense 7B specialized in code and reasoning, 128k context, ~4.1 GB VRAM in Q4. Lightweight, runs on a 6–8 GB GPU, MIT license.
ollama pull glm-5.3
Granite 4.2 8B (IBM): dense Apache 2.0, 128k context, ~4.6 GB Q4 VRAM. Multilingual chat, coding, and reasoning for the enterprise.
ollama pull granite4.2
SFT “thinking” fine-tune of OLMo 3 7B: step-by-step reasoning, 16k context, ~4.2 GB VRAM in Q4. 100% open, Apache 2.0 license.
# HuggingFace: zimplex/olmo3-7b-think-sft-eosfix-16k-3ep-euc
Dense 14B with hybrid thinking. Equals Qwen 2.5 32B Based on STEM/code.
ollama run qwen3:14b
Distilled R1 Qwen 14B. AIME24 69.7, MATH-500 93.9. Outperforms o1-mini on many benchmarks.
ollama run deepseek-r1:14b
Exceptional reasoning for its size. STEM-focused.
ollama run phi4:14b
| Rank | Model | Params | Q4 VRAM | Context | License | On RTX 4070 Super |
|---|---|---|---|---|---|---|
| #1 | Gemma 4 12B | 12B | 7 GB | 262 144 | Apache 2.0 | 28 tok/s · Q5_K_M |
| #2 | GLM 5.3 7B | 7B | 4.1 GB | 128 000 | MIT | 50 tok/s · Q8 |
| #3 | Granite 4.2 8B | 8B | 4.6 GB | 128 000 | Apache 2.0 | 50 tok/s · Q8 |
| #4 | OLMo 3 7B Think (SFT) | 7B | 4.2 GB | 16 000 | Apache 2.0 | 50 tok/s · Q8 |
| #5 | Qwen 3 14B | 14B | 9 GB | 131 072 | Apache 2.0 | 20 tok/s · Q5_K_M |
| #6 | DeepSeek R1 Distill Qwen 14B | 14B | 9 GB | 131 072 | MIT | 20 tok/s · Q5_K_M |
| #7 | Phi-4 14B | 14B | 9 GB | 16 384 | MIT | 20 tok/s · Q5_K_M |
Your private, free ChatGPT on your machine in 1 hour — LM Studio, Ollama, Open WebUI, your documents, no cloud.
Free memo
Get the memo VRAM → best coding model → Ollama command (one screen, copy and paste). Then switch to the Copilote Local kit for a setup that actually works.
The Local Copilot kit — the Ollama + Cline + Aider configs are ready to paste, with tuned Modelfiles, troubleshooting, and lifetime online access →No spam. Unsubscribe in 1 click. Your data stays with us (never resold).
Your card → the best coding model to run locally, and the exact Ollama command:
| Your VRAM | Typical GPUs / Macs | Recommended coding model | Ollama command |
|---|---|---|---|
| 8 GB | RTX 4060 / 3060 · M1-M2 16 GB | Qwen 3.5 9B (Q4, 6.6 GB — 256k context) | ollama run qwen3.5:9b |
| 12 GB | RTX 3060 12 GB / 4070 / 5070 | Qwen 3.5 9B (Q8, 11 GB) or Gemma 4 12B (7.6 GB) | ollama run qwen3.5:9b-q8_0 |
| 16 GB | RTX 5070 Ti / 4080 / 5080 · RX 9070 XT · M4 24 GB | Devstral 24B (Q4, 14 GB) — coding-agent specialist | ollama run devstral:24b |
| 24 GB | RTX 3090 / 4090 · RX 7900 XTX · M4 Pro 48 GB | Qwen 3.8 27B (Q4, 18 GB) — the “close to Copilot” option | ollama run qwen3.8:27b |
| 32 GB | RTX 5090 | Qwen 3.6 35B-A3B (Q4, 23 GB) — fast MoE | ollama run qwen3.6:35b |
| 48 GB+ | Mac M4 Max 64 GB · M2 Ultra 128 GB | Qwen3-Coder 30B-A3B (Q8, 32 GB — 256k context) | ollama run qwen3-coder:30b-a3b-q8_0 |
-base: ollama run qwen2.5-coder:7b-base — it’s still the reference for this specific use case. ⚠️ Qwen 3.8: its reasoning is set very high by default and it “overthinks” simple requests — lower it to low (or turn it off) on first launch. ⚠️ License trap: Codestral 22B = Mistral Non-Production License → prohibited for coding at work. Qwen 3.5/3.8, Gemma 4, and Devstral are Apache 2.0. 💡 Running out of memory? Keep ~1.5 GB of VRAM free for context, or drop down one quantization level.🔌 To connect it to VS Code: Cline (multi-file agent), Aider (CLI) or Tabby/Twinny (FIM autocomplete) — they all connect to Ollama locally. The kit Local Copilot — ready-to-paste configs + tested setup — is available: /copilote-local.
Filter: Q4_K_M ≤ 11 GB. 7-14B bonus. ~15% boost vs. standard 4070 variant.
Criteria considered:
The scoring is fully transparent: see our methodology for details on VRAM/tokens/sec calculations.
4070 Super vs. 4070?
Even 12 GB. 4070 Super = +15% CUDA cores and bandwidth. Mistral 7B Q4: ~35 tok/s vs ~30 tok/s. See RTX 4070.
4070 Super or 5070?
5070 = ~25% faster (GDDR7) but ~€150 more. Excellent 4070 Super used for ~€450. See RTX 5070.
LLM sweet spot for the 4070 Super?
Qwen 3 14B Q4 or Mistral Nemo 12B Q5. 35–45 tok/s. Excellent for personal RAG (32k context fits).
Can Mistral Small 24B run?
In Q3_K_M (~10 GB), yes, but with degraded quality. Q4 (~13 GB) does not fit. For comfortable 24B performance, target a 4070 Ti Super 16 GB. See 4070 Ti Super.
Prices in euros (€) are French market prices including VAT, as checked by BestLLMfor. US prices differ: the Amazon buttons show the current US price.