Instructions to use moonshotai/Kimi-K3.1-Pro with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moonshotai/Kimi-K3.1-Pro with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="moonshotai/Kimi-K3.1-Pro", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForSeq2SeqLM
model = AutoModelForSeq2SeqLM.from_pretrained("moonshotai/Kimi-K3.1-Pro", trust_remote_code=True, device_map="auto")

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use moonshotai/Kimi-K3.1-Pro with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moonshotai/Kimi-K3.1-Pro"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moonshotai/Kimi-K3.1-Pro",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/moonshotai/Kimi-K3.1-Pro

SGLang

How to use moonshotai/Kimi-K3.1-Pro with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "moonshotai/Kimi-K3.1-Pro" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moonshotai/Kimi-K3.1-Pro",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "moonshotai/Kimi-K3.1-Pro" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moonshotai/Kimi-K3.1-Pro",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use moonshotai/Kimi-K3.1-Pro with Docker Model Runner:
```
docker model run hf.co/moonshotai/Kimi-K3.1-Pro
```

Add evaluation results

by SaylorTwift HF Staff - opened about 5 hours ago

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

about 5 hours ago

Add evaluation results for moonshotai/Kimi-K3.1-Pro

Summary

This PR adds evaluation results from the model card's benchmark table for moonshotai/Kimi-K3.1-Pro to the .eval_results/ directory, following the Hugging Face Hub evaluation-results specification.

Benchmarks Added

Terminal-Bench 4.0: 65.7
Terminal-Bench-Science 0.1: 38.1
DeepSWE v1.1: 75.5
HLE-Full (no tools / w/ tools): 51.1 / 70.0
GPQA-Diamond: 94.2
MMMU-Pro (plain / w/ python): 84.1 / 90.0
Agents' Last Exam (ALL): 43.2
Agents' Last Exam (CLI): 44.2
APEX-Agents: 69.4

Mapping notes:

MMMU-Pro → mmmu_pro_vision. The card doesn't name the setting; this follows how earlier Kimi releases (K2.5, K2.6) are listed on this leaderboard.
Agents' Last Exam → the pass-rate tasks. The card doesn't say whether it reports pass rate or score.
HLE and MMMU-Pro each have two entries (with and without tools), separated by notes.

Benchmarks Skipped (Not Registered on Hub)

The following benchmarks are in the model card but have no registered eval.yaml on the Hugging Face Hub:

SWE Marathon v1.1: 51.7
BenchCAD (w/o tools / w/ tools): 72.8 / 92.7
ProgramBench: 31.5 (dataset exists but has no eval.yaml yet)
PostTrainBench v1.2: 36.5
BrowseComp: 92.1
BioMysteryBench: 80.6
CharXiv (RQ) (plain / w/ python): 89.0 / 93.4
GDP.pdf (plain / w/ python): 34.6 / 85.7
ChartoGraphy w/ python: 90.2
BabyVision w/ python: 95.4
ZeroBench_main (avg@5 / pass@5; w/ python pass@5): 29.8 / 39.0; 63.0
AutomationBench-AA: 73.3
GDPval-AA v2.1 (Elo): 1846
AA-Briefcase v1.1 (Elo): 1678
OfficeQA Pro v2 / OfficeQA Pro: 64.4 / 67.4
Frontier Finance: 58.6
Harvey Legal Agent: 23.7
HANDBOOK.md: 44.6
OSWorld V2.1: 79.6
HLE-Diamond: 39.4 (not a task of cais/hle)
ExploitBench: 74.7
ExploitGym: 20.7

Internal benchmarks (not public, so not eligible): Kimi Code Bench 2.0 (77.1), Kimi Code Bench 3.0 Long-horizon (61.2), Business Operation (37.8), DECK-Arena (1456 Elo), Design Bench (957 Elo), PerceptionBench (68.3), Kimi Hallu Bench Hard w/ tools (8.3) and w/o tools (5.4), xdev-bench-hard (81.0).

These can be added once the benchmark authors register an eval.yaml on the Hub.

Source

Model card: https://huggingface.co/moonshotai/Kimi-K3.1-Pro

Files Added

.eval_results/Kimi-K3.1-Pro.yaml

Verification

These results come from the model card's published benchmark table: reasoning effort max, temperature = 1.0, top-p = 0.95 unless the card says otherwise. No verified token is provided, since these were not run via HF Jobs with inspect-ai.

Add evaluation results9189f397

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

Ready to merge

This branch is ready to get merged automatically.

· Sign up or log in to comment