NVIDIA · 10GB VRAM
What AI models can a GeForce RTX 3080 run?
A GeForce RTX 3080 has 10GB of VRAM. Checked against 228 models from the Ollama library, 156 run fully on the GPU at 4K context and 44 more run with part of the model in system RAM.
156
run well
44
with trade-offs
228
models tested
Best AI models for a GeForce RTX 3080
Largest first — these fit entirely in 10GB of VRAM at 4K context, so the whole model is GPU-accelerated.
| Model | Build | Memory needed | Verdict |
|---|---|---|---|
| nous-hermes2Ollama Library | 10.7B | ~9.0 GB | Good fit |
| solarUpstage | 10.7B | ~9.0 GB | Good fit |
| codegeex4Zhipu AI | 9B | ~8.7 GB | Good fit |
| glm4Zhipu AI | 9B | ~8.7 GB | Good fit |
| ornith-1.5Ollama Library | 9B | ~7.9 GB | Good fit |
| ornithOllama Library | 9B | ~7.9 GB | Good fit |
| gemma3nGoogle DeepMind | e2b-it-q4_K_M | ~7.9 GB | Good fit |
| lfm2.5Ollama Library | 8B | ~7.3 GB | Excellent fit |
| granite4.1-guardianIBM | 8B | ~7.2 GB | Excellent fit |
| rnj-1Ollama Library | 8B | ~7.2 GB | Excellent fit |
| command-r7b-arabicCohere | 7B | ~7.1 GB | Excellent fit |
| command-r7bCohere | 7B | ~7.1 GB | Excellent fit |
| aya-expanseCohere | 8B | ~7.1 GB | Excellent fit |
| llama-proMeta AI | 8B | ~7.1 GB | Excellent fit |
| minicpm-v4.5Ollama Library | 8B | ~7.1 GB | Excellent fit |
| llava-llama3Ollama Library | 8B | ~6.9 GB | Excellent fit |
| tulu3Allen Institute for AI | 8B | ~6.9 GB | Excellent fit |
| llama3-groq-tool-useMeta AI | 8B | ~6.9 GB | Excellent fit |
| dolphin3Ollama Library | 8B | ~6.9 GB | Excellent fit |
| dolphin-llama3Ollama Library | 8B | ~6.9 GB | Excellent fit |
| llama3.1Meta AI | 8B | ~6.9 GB | Excellent fit |
| llama3-gradientMeta AI | 8B | ~6.9 GB | Excellent fit |
| llama3Meta AI | 8B | ~6.9 GB | Excellent fit |
| llama3-chatqaMeta AI | 8B | ~6.9 GB | Excellent fit |
| ayaCohere | 8B | ~6.8 GB | Excellent fit |
Showing the 25 largest of 156 models that run well.
How these results were calculated
Each model is evaluated at 4K context against 10GB of VRAM, counting model weights, KV cache, compute buffers and runtime overhead, minus a reserve for the display and operating system. Download sizes come from the official Ollama registry.
These figures assume 32GB of system RAM and working GPU drivers. Your own machine may differ — RAM, free disk space and whether an acceleration backend is actually installed all change the answer, which is what the PC scan measures directly.
GeForce RTX 3080 local AI — frequently asked questions
- How many AI models can a GeForce RTX 3080 run?
- Out of 228 models in the Ollama library, 156 run fully on a GeForce RTX 3080's 10GB of VRAM at 4K context, and a further 44 run with part of the model offloaded to system RAM, more slowly.
- What is the largest AI model a GeForce RTX 3080 can run?
- The heaviest build that fits entirely in VRAM is nous-hermes2:10.7b-solar-q4_K_M (10.7B parameters), needing about 9.0 GB of memory from a 6.0 GB download.
- Is 10GB of VRAM enough for local AI?
- 10GB is enough for 156 of the 228 models tested here, which covers most general-purpose and coding assistants. Very large models still need either partial CPU offload or a card with more memory.