can i run · seo
Yes — InternVL3 8B runs on RTX 4090 24GB
InternVL3 8B (8B) needs 16GB VRAM at FP16 vs your 24GB VRAM. Estimated ~36–49 tok/s.
Strong open vision-language at 8B. Q4 ~6 GB. Screenshot / document QA.
best fit on this rig · weights + KV + headroom
FP16 · 16GB weights+0.4GB KV @ 8kfits GPU~36–49 tok/s
| quant | weights | + KV 8k | fits 24GB? |
|---|---|---|---|
| Q4 | 6GB | ~6.4GB | ✓ yes |
| Q5 | 7.5GB | ~7.9GB | ✓ yes |
| Q8 | 10GB | ~10.4GB | ✓ yes |
| FP16 | 16GB | ~16.4GB | ✓ yes |
File size ≠ runtime. KV grows with context — halve to 4k/2k if long chats OOM. Leave 1–2GB free.
llama-cli -hf unsloth/InternVL3-8B-GGUF:Q4_K_M --mmproj auto
Run it: 1) install Ollama · 2) copy command · 3) verify with ollama ps (want 100% GPU).
faq · straight answers
Can InternVL3 8B run on RTX 4090 24GB?
InternVL3 8B (8B) needs 16GB VRAM at FP16 vs your 24GB VRAM. Estimated ~36–49 tok/s. Model context 32k. Compare with all models or re-check with your exact specs on the bench.
What if it OOMs?
Unload with ollama stop, cap layers --num-gpu 28, cut context to 2k. See the OOM fixer on the homepage.
Q4 or Q8 on this rig?
Start FP16. Bigger model at Q4 beats smaller at Q8. Only go Q8 if 2GB+ headroom remains.
How fast will it be?
About ~36–49 tok/s when 100% on GPU. Any CPU offload drops speed 5–20× — check ollama ps.
Is Apache-2.0 commercial-safe?
Yes — Apache-2.0 allows commercial use.
also fits RTX 4090 24GB
InternVL3 8B on other rigs