FreeLLMGPUs › NVIDIA
Best free AI models for the RTX 4090
RTX 4090: 24 GB of video memory, 1008 GB/s. These are the free models it runs, ranked by Arena votes, with the speed you can expect.
Simulate your setup · Check the price on Amazon ↗ · Compare all GPUs
Voted models that fit fully on the GPU
| Model | Arena score | Size | 4-bit file | Tokens/s |
|---|---|---|---|---|
| Gemma 4 31B |
1443 | 31.3 B | 20 GB | 36 |
| Qwen3.8-27B alibaba |
1441 | 27.8 B | 18 GB | 43 |
| Gemma 4 26B A4B google · MoE, 4 B active |
1434 | 25.8 B | 17 GB | 152 |
| Qwen3.5-27B alibaba |
1409 | 27.8 B | 18 GB | 43 |
| gemma-3-27b-it |
1358 | 27.4 B | 18 GB | 42 |
Bigger models with help from 64 GB of system RAM
Part of the model sits in system memory, which is much slower than video memory.
Popular small models
Official small models with no Arena votes yet, ordered by downloads on Hugging Face (popularity, not quality).
| Model | Downloads/month | Size | 4-bit file | Tokens/s |
|---|---|---|---|---|
| Qwen3-0.6B Qwen |
29.5M | 0.8 B | 1.7 GB | 356 |
| Qwen3-4B Qwen |
11.8M | 4 B | 3.6 GB | 178 |
| Qwen3-8B Qwen |
8.9M | 8.2 B | 6.2 GB | 113 |
| Qwen3-14B Qwen |
8.4M | 14.8 B | 10 GB | 71 |
| Qwen3.5-4B Qwen |
7.8M | 4.7 B | 4.0 GB | 201 |
| Qwen3.5-9B Qwen |
7.5M | 9.7 B | 7.1 GB | 113 |
Estimates at 4-bit (Q4_K_M) with an 8K context, calibrated on 44 real llama.cpp tests (typical error about ±10%). Tokens per second: about 10 is reading speed. As an Amazon Associate, FreeLLMArena earns from qualifying purchases; the links do not change the rankings.
Questions
What is the best free AI model for the RTX 4090?
Gemma 4 31B (Arena score 1443) is the highest-voted open model that fits entirely in 24 GB of video memory at 4-bit, writing about 36 tokens per second.
How big a model can the RTX 4090 run?
About 35 billion parameters for a dense model at 4-bit with an 8K context, fully on the GPU. Mixture-of-experts models can be bigger, and with 64 GB of system RAM larger models still run, more slowly.
Is the RTX 4090 good for local LLMs?
It has 24 GB of memory and 1008 GB/s of bandwidth. Memory decides which models fit; bandwidth decides how fast they write. With 24 GB or more it runs 27–32 B models, the sweet spot of quality for home use.
Similar cards
- RTX 3090 Ti 24 GB · 1008 GB/s
- RX 7900 XTX 24 GB · 960 GB/s
- RTX 3090 24 GB · 936 GB/s
- RX 7900 XT 20 GB · 800 GB/s
- RTX 5090 32 GB · 1792 GB/s
- RTX 5080 16 GB · 960 GB/s