InternLM 3 — local AI model by Shanghai AI Lab
Written by Jakub Rusinowski · Last updated
A powerful open-source model from China's Shanghai AI Lab. InternLM 3 excels at bilingual Chinese-English tasks, coding, and long document analysis. Particularly strong in math and STEM reasoning, making it the top open choice for Chinese language applications.
Variants
The smallest InternLM 3 variant needs about 6 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| InternLM 3 8B Instruct → 8B | ~6.1 GB | 32,768 | ollama run internlm3:8b | |
| InternLM 3 20B Instruct → 20B | ~12.9 GB | 32,768 | ollama run internlm3:20b |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run InternLM 3 locally
Install Ollama, then pull the tag.
ollama run internlm3:8bPick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: InternLM 3 8B Instruct, InternLM 3 20B InstructRecommended GPU
The cheapest catalogued GPU that runs InternLM 3 locally (min 6 GB VRAM) is the Intel Arc B570 (10 GB).
Can I run InternLM 3 on my GPU?
- InternLM 3 on AMD Radeon RX 7800 XT
- InternLM 3 on AMD Radeon RX 7900 XT
- InternLM 3 on AMD Radeon RX 7900 XTX
- InternLM 3 on AMD Radeon RX 9060 XT 16GB
- InternLM 3 on AMD Radeon RX 9060 XT 8GB
- InternLM 3 on AMD Radeon RX 9070
- InternLM 3 on AMD Radeon RX 9070 XT
- InternLM 3 on Apple M2
- InternLM 3 on Apple M3
- InternLM 3 on Apple M4 Pro
- InternLM 3 on Intel Arc B570
- InternLM 3 on Intel Arc B580
- InternLM 3 on NVIDIA GeForce RTX 3060 (12GB)
- InternLM 3 on NVIDIA GeForce RTX 3070
- InternLM 3 on NVIDIA GeForce RTX 3070 Ti
- InternLM 3 on NVIDIA GeForce RTX 3080 (10GB)
- InternLM 3 on NVIDIA GeForce RTX 3090
- InternLM 3 on NVIDIA GeForce RTX 4060
- InternLM 3 on NVIDIA GeForce RTX 4060 Ti 16GB
- InternLM 3 on NVIDIA GeForce RTX 4070
- InternLM 3 on NVIDIA GeForce RTX 4070 Super
- InternLM 3 on NVIDIA GeForce RTX 4070 Ti
- InternLM 3 on NVIDIA GeForce RTX 4070 Ti Super
- InternLM 3 on NVIDIA GeForce RTX 4080
- InternLM 3 on NVIDIA GeForce RTX 4080 Super
- InternLM 3 on NVIDIA GeForce RTX 4090
- InternLM 3 on NVIDIA GeForce RTX 5060
- InternLM 3 on NVIDIA GeForce RTX 5060 Ti 16GB
- InternLM 3 on NVIDIA GeForce RTX 5060 Ti 8GB
- InternLM 3 on NVIDIA GeForce RTX 5070
- InternLM 3 on NVIDIA GeForce RTX 5070 Ti
- InternLM 3 on NVIDIA GeForce RTX 5080
InternLM 3 — frequently asked questions
How much VRAM does InternLM 3 need?
InternLM 3 needs about 6 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: InternLM 3 8B Instruct (6 GB, Q4_K_M); InternLM 3 20B Instruct (13 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run InternLM 3 on an RTX 4090 (24 GB)?
Yes — InternLM 3 runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for InternLM 3?
Q4_K_M is the best balance of quality and VRAM for InternLM 3 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run InternLM 3 with Ollama?
Install Ollama, then run: ollama run internlm3:8b. This downloads InternLM 3 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.