InternVL2-Llama3-76B
76B multimodal model pairing InternViT-6B vision with Hermes-2-Theta Llama-3-70B.
About
InternVL2-Llama3-76B is the largest member of the InternVL 2.0 series from OpenGVLab, released in July 2024. It combines the InternViT-6B-448px-V1-5 vision encoder with the Hermes-2-Theta-Llama-3-70B language model, totaling 76 billion parameters.
Like the rest of the series it is instruction-tuned, trains on an 8k context window with long texts, multiple images and video data, and targets document and chart comprehension, OCR, scientific and mathematical problem solving, and cultural understanding. It is distributed as open weights through Hugging Face.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 123; Trending score 64/100 from researcher assessment.