SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.16855 · 2026

Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agents

Ming Yan, Xi Zhang, Junyang Lin, Jingren Zhou, Zhiyuan Chen, Junyang Wang, Haiyang Xu, Shuai Bai, Junjie Cao, Xuhao Hu, Haowei Liu, Feiyu Gao, and 7 more

arXiv · PDF · Open in the Atlas

Code that ran

6 of 11 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 5 failures of our instrument; 5 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2602.16855.json · badge: /api/ran/2602.16855.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
X-PLUG/MobileAgent canonical (classifier label)
classifier basis: census
6 ran · 5 no recorded run
FunctionStatusWhere it lives
calculate_iou Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/crop.py
code served (permissive licence) · get_code("104b4a02b11c457a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# box1 and box2 are expected to be sequences/lists with 4 elements
# representing [x1, y1, x2, y2] coordinates
# The function uses indexing with [0], [1], [2], [3]

box1 = [10, 20, 50, 60]
box2 = [30, 40, 70, 80]

ARGS = {
    "box1": box1,
    "box2": box2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
calculate_size Ran
ran, violated the contract we drafted
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/crop.py
code served (permissive licence) · get_code("53fa5e1f48e2eea0")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int, ndim 0, dtype family int, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# box is expected to be a sequence with at least 4 elements
# box[0], box[1], box[2], box[3] are used to calculate area
# The function treats them as: (x1, y1, x2, y2) coordinates
box = [10, 20, 50, 80]

ARGS = {
    "box": box
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
crop_image Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/crop.py
code served (permissive licence) · get_code("b371593bf6e485b9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family int, shape (60, 80, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import cv2

# Create a small test image (100x100 RGB image)
img = np.random.randint(0, 256, (100, 100, 3), dtype=np.uint8)

# Create position as a numpy array with 4 corner points
# These should form a quadrilateral that can be perspective-transformed
position = np.array([
    [10, 20],      # top-left
    [90, 25],      # top-right
    [95, 85],      # bottom-right
    [15, 80]       # bottom-left
], dtype=np.float32)

ARGS = {
    "img": img,
    "position": position
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (60, 80, 3) — shape (60, 80, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
encode_image Ran
ran; our contract draft was wrong, not the code
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/api.py
pointer only (licence: MIT); the stored copy get_code serves, from jongwoopark7978/LVNet, is pointer only (licence: NONE) · get_code("f41cb1a19b154297")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from jongwoopark7978/LVNet, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

init_chat Ran
ran; our contract draft was wrong, not the code
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/chat.py
code served (permissive licence) · get_code("303ea64d50ae9fbb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

ARGS = {
    'instruction': 'Please unlock my phone and open the camera app'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
order_point Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/text_localization.py
code served (permissive licence) · get_code("4a27a65acf4c80a5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (4, 2) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# The function order_point expects a coordinate input that can be reshaped to [4, 2]
# This means it needs 8 elements total (4 points with 2 coordinates each)

# Create a simple set of 4 points (8 coordinates)
coor = [0.0, 0.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0]

ARGS = {
    "coor": coor
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (4, 2) — shape (4, 2) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_action Not yet run
unverified (harvested, no recorded run)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/api_service.py
code served (permissive licence) · get_code("5434906159fc3994")
How it ran

no invocation recorded.

get_size Not yet run
unverified (harvested, no recorded run)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/controller.py
code served (permissive licence) · get_code("f244ecf0b63d34d8")
How it ran

no invocation recorded.

inference_chat Not yet run
unverified (harvested, no recorded run)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/api.py
code served (permissive licence) · get_code("c03862724ffdab8d")
How it ran

no invocation recorded.

inference_chat Not yet run
unverified (harvested, no recorded run)
X-PLUG/MobileAgent/Mobile-Agent-v2/MobileAgent/api.py
code served (permissive licence) · get_code("b25b6801b4b7729c")
How it ran

no invocation recorded.

longest_common_substring_length Not yet run
unverified (harvested, no recorded run)
X-PLUG/MobileAgent/Mobile-Agent-v1/MobileAgent/text_localization.py
code served (permissive licence) · get_code("63e598cdc52304ad")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 1 the same month, 1 later) and 10 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
encode_image 76 other papers
74 dated earlier, 1 the same month, 1 later
2309.05950 (2023-09)
Language Models as Black-Box Optimizers for Vision-Language Models
2311.15732 (2023-11)
GPT4Vis: What Can GPT-4 Do for Zero-shot Visual Recognition?
2312.14867 (2023-12)
VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluation
and 73 more not named here

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The paper introduces GUI-Owl-1.5, the latest native GUI agent model that features instruct/thinking variants in multiple sizes (2B/4B/8B/32B/235B) and supports a range of platforms (desktop, mobile, browser, and more) to enable cloud-edge collaboration and real-time interaction. GUI-Owl-1.5 achieves state-of-the-art results on more than 20+ GUI benchmarks on open-source models: (1) on GUI automation tasks, it obtains 56.5 on OSWorld, 71.6 on An-droidWorld, and 48.4 on WebArena; (2) on grounding tasks, it obtains 80.3 on ScreenSpotPro; (3) on tool-calling tasks, it obtains 47.6 on OSWorld-MCP, and 46.8 on MobileWorld; (4) on memory and knowledge tasks, it obtains 75.5 on GUI-Knowledge Bench. GUI-Owl-1.5 incorporates several key innovations: (1) Hybird Data Flywheel: we construct the data pipeline for UI understanding and trajectory generation based on a combination of simulated environments and cloud-based sandbox environments, in order to improve the efficiency and quality of data collection. ( 2) Unified Enhancement of Agent Capabilities: we use a unified thought-synthesis pipeline to enhance the model's reasoning capabilities, while placing particular emphasis on improving key agent abilities, including Tool/MCP use, memory and multi-agent adaptation; (3) Multi-platform Environment RL Scaling: We propose a new environment RL algorithm, MRPO, to address the challenges of multi-platform conflicts and the low training efficiency of long-horizon tasks. The GUI-Owl-1.5 models are open-sourced, and an online cloud-sandbox demo is available at https://github.com/X-PLUG/MobileAgent.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2602.16855")
get_code_for_paper("2602.16855")
have("2602.16855")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2602.16855.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2602.16855.svg)](https://syntology.ai/paper/2602.16855)

Connect an agent — have() is free.