GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
x-square-robot
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
x-square-robot
Model Tree
Input Modalities
Output Modalities
Supported Functionality
| Property | Value |
|---|---|
| Release | X-Planner-9B-0916 |
| Architecture | Qwen3_5ForConditionalGeneration (Qwen3.5 9B architecture) |
| Stored parameters | 9,409,813,744 |
| Weight precision | BF16 |
| Serialization | Safetensors, sharded at 5 GB |
| Transformers version recorded by the checkpoint | 5.2.0 |
The repository includes all model weights, the model/generation configuration,
tokenizer, chat template, image/video processor configuration, and minimal
checkpoint metadata. Sharding preserves all 760 tensors bit for bit.
release_manifest.json records file hashes and the original, unsharded weights
SHA-256 for provenance. Training logs, optimizer state, and machine-specific
training paths are not required for inference and are not part of this release.
bash
python -m pip install -U huggingface_hubhf download x-square-robot/X-Planner-9B-0916 \--local-dir checkpoints/X-Planner-9B-0916
For reproducible runs, add --revision <commit> using the desired revision from
this repository's commit history.
Use Transformers with Qwen3.5 support (the checkpoint was saved with 5.2.0), PyTorch, and Accelerate. No custom remote model code is required.
python
import torchfrom transformers import AutoModelForImageTextToText, AutoProcessorcheckpoint = "checkpoints/X-Planner-9B-0916"processor = AutoProcessor.from_pretrained(checkpoint)model = AutoModelForImageTextToText.from_pretrained(checkpoint,dtype=torch.bfloat16,device_map="auto",attn_implementation="sdpa",).eval()model.config.use_cache = Truemodel.config.text_config.use_cache = True
The weights occupy approximately 18.82 GB in BF16. Inference also needs memory for activations, visual tokens, and the generation cache; total memory use depends on the inputs and generation length.
The task-specific prompt, image preparation, history format, and output parser are defined by the X-Planner event-state runtime. After installing its dependencies and preparing an event snapshot, run from the code checkout:
bash
python scripts/inference/run_event_planner.py \--checkpoint checkpoints/X-Planner-9B-0916 \--snapshot /path/to/event_snapshot \--output-dir work_dirs/inference
The event-state CLI requires an X-Planner-compatible data backend and a prepared event snapshot. See the code repository's installation and data-preparation instructions for the backend's public availability. The benchmark's raw video manifest is not a prepared event snapshot.
The published benchmark contains 1,500 episodes, 3,490 videos, and episode-level planning metadata, with playable multi-view videos in Dataset Preview. Full temporal scoring annotations and a fixed end-to-end evaluation protocol are separate from this media release.
This model card does not report a new evaluation of X-Planner-9B-0916. Results from the technical report or other checkpoint revisions should retain their original model and evaluation provenance.
Model weights are distributed under Apache 2.0, consistent with the Qwen3.5-9B architecture's upstream model release. The X-Planner source code is MIT-licensed. Benchmark data and media retain their respective upstream terms, as described in the dataset card.
bibtex
@article{xplanner2026event,title = {X-Planner: Event-Structured Task Planning for Embodied Intelligence},author = {{X Square Robot Team}},year = {2026},note = {Technical report}}