565
Total profiles
Structured model pages for long-tail SEO: grouped by license scope, model size, and workload scenario. Every profile includes VRAM requirements, local-vs-cloud recommendation, and benchmark anchor metrics.
Ollama naming/source alignment date: 2026-02-24. Each profile links back to its Ollama library source.
Total profiles
Group hubs
Measured profiles
Estimated profiles
SC keywords recycled
Top curated models
| Rank | Model | Tag | VRAM Q4 | VRAM Q5 | Category | Command |
|---|---|---|---|---|---|---|
| 1 | Llama 3 | llama3:8b | 10GB | 12GB | General Chat | ollama run llama3:8b |
| 2 | Llama 3 | llama3:70b | 24GB | 30GB | General Chat | ollama run llama3:70b |
| 3 | Llama 3.3 | llama3.3:70b | 24GB | 30GB | General Chat | ollama run llama3.3:70b |
| 4 | Mistral | mistral:7b | 8GB | 10GB | General Chat | ollama run mistral:7b |
| 5 | Mixtral | mixtral:8x7b | 20GB | 24GB | General Chat | ollama run mixtral:8x7b |
| 6 | Gemma 2 | gemma2:9b | 10GB | 12GB | General Chat | ollama run gemma2:9b |
| 7 | Gemma 2 | gemma2:27b | 16GB | 20GB | General Chat | ollama run gemma2:27b |
| 8 | DeepSeek-R1 | deepseek-r1:67b | 24GB | 30GB | Reasoning / Math | ollama run deepseek-r1:67b |
| 9 | DeepSeek-V3 | deepseek-v3:67b | 24GB | 30GB | General / Reasoning | ollama run deepseek-v3:67b |
| 10 | Zephyr | zephyr:7b | 8GB | 10GB | Assistant Chat | ollama run zephyr:7b |
| 11 | OpenHermes | openhermes:7b | 8GB | 10GB | Assistant Chat | ollama run openhermes:7b |
| 12 | CodeLlama | codellama:7b | 8GB | 10GB | Coding | ollama run codellama:7b |
| 13 | CodeLlama | codellama:34b | 16GB | 20GB | Coding | ollama run codellama:34b |
| 14 | DeepSeek Coder | deepseek-coder:33b | 16GB | 20GB | Coding | ollama run deepseek-coder:33b |
| 15 | Qwen2.5 Coder | qwen2.5-coder:32b | 16GB | 20GB | Coding | ollama run qwen2.5-coder:32b |
| 16 | Llama 3.2 Vision | llama3.2-vision:11b | 12GB | 14GB | Multimodal / Vision | ollama run llama3.2-vision:11b |
| 17 | LLaVA | llava:7b | 8GB | 10GB | Multimodal / Vision | ollama run llava:7b |
| 18 | Phi-3 | phi3:3.8b | 4GB | 5GB | Lightweight / Entry | ollama run phi3:3.8b |
| 19 | Gemma 2 | gemma2:2b | 2GB | 3GB | Lightweight / Entry | ollama run gemma2:2b |
| 20 | TinyLlama | tinyllama:1.1b | 2GB | 3GB | Lightweight / Entry | ollama run tinyllama:1.1b |
84 models
Open group471 models
Open group10 models
Open group196 models
Open group150 models
Open group10 models
Open group111 models
Open group98 models
Open group34 models
Open group72 models
Open group26 models
Open group100 models
Open group53 models
Open group56 models
Open group109 models
Open group115 models
Open group| Model | VRAM (min/optimal) | 3090 tok/s | Path |
|---|---|---|---|
| Llama 4 128X17B Q4 | 418GB / 428GB | 1.1 | View |
| Llama 4 128X17B Q5 | 420GB / 430GB | 1 | View |
| Llama 4 128X17B Q8 | 424GB / 434GB | 0.8 | View |
| Llama 4 128X17B FP16 | 430GB / 442GB | 0.6 | View |
| DeepSeek-R1 671B Q4 | 418GB / 428GB | 1.1 | View |
| DeepSeek-R1 671B Q5 | 420GB / 430GB | 1 | View |
| DeepSeek-R1 671B Q8 | 424GB / 434GB | 0.8 | View |
| DeepSeek-R1 671B FP16 | 430GB / 442GB | 0.6 | View |
Top keywords feeding daily content and model page updates.
| Keyword | Clicks | Impressions | CTR | Landing |
|---|---|---|---|---|
| llama 70b on 3090 | 31 | 702 | 4.4% | /en/models/llama-70b-q4/ |
| q4 vs q8 quality ollama | 22 | 581 | 3.8% | /en/tools/quantization-blind-test/ |
| best local rag model | 17 | 429 | 4% | /en/blog/best-local-rag-models-2026/ |
| runpod a100 ollama | 14 | 191 | 7.3% | /en/affiliate/cloud-gpu/ |