Llama 3.1 8B vs Llama 3.3 70B
Comparing VRAM requirements, performance, and capabilities for running these models locally with Ollama.
Parameters
8B
Context
128K
VRAM Range
6.3–18 GB
Recommended
Q8_0 (10 GB)
ByMeta·LicenseLlama 3.1 Community License
Parameters
70B
Context
128K
VRAM Range
43.5–72 GB
Recommended
Q4_K_M (43.5 GB)
ByMeta·LicenseLlama 3.3 Community License
VRAM Requirements by Quantization
Side-by-side memory needs at each quality level.
| Quantization | Llama 3.1 8B | Llama 3.3 70B | Difference |
|---|---|---|---|
| Q4_K_M | 6.3 GB | 43.5 GB | -37.2 GB |
| Q8_0 | 10 GB | 72 GB | -62.0 GB |
| F16 | 18 GB | — | — |
Capabilities
Feature support comparison.
| Capability | Llama 3.1 8B | Llama 3.3 70B |
|---|---|---|
| text generation | Yes | Yes |
| code generation | Yes | Yes |
| multilingual | Yes | Yes |
| tool use | Yes | Yes |
| summarization | Yes | Yes |
| reasoning | — | Yes |
| math | — | Yes |
| creative writing | — | Yes |
Benchmark Scores
Higher is better. Scores from published evaluations.
| Benchmark | Llama 3.1 8B | Llama 3.3 70B |
|---|---|---|
| mmlu | 73.0 | 86.0 |
Hardware Compatibility
Can each model run at recommended quantization on common VRAM tiers?
| VRAM | Llama 3.1 8B | Llama 3.3 70B |
|---|---|---|
| 8 GB | Offload | No |
| 12 GB | Runs | No |
| 16 GB | Runs | No |
| 24 GB | Runs | No |
| 32 GB | Runs | Offload |
| 48 GB | Runs | Tight |
| 64 GB | Runs | Runs |
| 96 GB | Runs | Runs |
Run Llama 3.1 8B
ollama run llama3.1:8b-instruct-q8_0Run Llama 3.3 70B
ollama run llama3.3:70b-instruct-q4_K_MCheck your exact hardware
Use the compatibility checker to see how each model performs on your specific GPU or Mac.