Vision-only weights (MoonViT tower + multimodal projector) extracted from
moonshotai/Kimi-K2.7-Code
for use with MLX-based inference stacks such as
exo,
in the same format as
exolabs/Kimi-K2.6-vision.
1capabilities = ["text", "thinking", "thinking_toggle", "vision"]
2
3[vision]
4image_token_id = 163605
5model_type = "kimi_vl"
6weights_repo = "aidiffuser/Kimi-K2.7-Code-vision"
7processor_repo = "moonshotai/Kimi-K2.7-Code"
Tested working: distributed (2× Mac Studio M3 Ultra, tensor parallelism) with
the official INT4 text weights, image understanding confirmed.
Same Modified MIT license as the source model; these are a subset of the
original weights, unmodified. All credit to Moonshot AI.
