← Back to directory
Shanghai AI Lab · Released 2026-01

InternLM 3

Shanghai AI Lab's third-generation InternLM. Apache-2.0, 200K context, strong reasoning. Quietly one of the best open Chinese-led models.

Apache-2.0Commercial use OKgeneralreasoning
Params (max)
70B
Variants
70B / 20B / 7B
Context window
200K tokens
MMLU
80.4
HumanEval
73
GSM8K
89.6
Min VRAM (fp16, smallest variant)
14GB
Smallest Q4 GGUF
~4GB
Languages supported
18
Pros
  • ✓Long context
  • ✓Permissive license
  • ✓Good multilingual
Cons
  • ×Less Western community traction
  • ×200K context heavy on memory

Highlights

  • ●200K context
  • ●Apache-2.0
  • ●Strong RAG performance

Where to download

Hugging Face: internlm/internlm3-70b-chat
Or via Ollama (ollama pull internlm-3) or LM Studio's in-app browser.
Homepage: https://internlm.org