RoboVLMs（KosMos P.H.）

Commercial use OK 7/8 verified

1 of 8 required fields are unverified. They are shown as badges, not filled with defaults. Sources for verified fields are listed below.

Facts

Released: 2024-12 1 Source
Parameters: Unverified 0 Source
Architecture: VLM + action expert 0 Source
License: Apache-2.0 (code and model card) 2 Source
Commercial: Commercial use OK 0 Source
Weights: huggingface.co ↗ 1 Source
Code: github.com ↗ 1 Source
Paper: arxiv.org ↗ 0 Source
Input modalities: rgb · language 0 Source
Action space: End-effector pose delta + gripper 0 Source
Cross-embodiment: Open X-Embodiment pretraining; CALVIN / SimplerEnv / real robot 0 Source
Pretraining data (hours): Unverified 0 Source
Target forms: Single arm 0 Source

Architecture · Framework for turning any VLM into a VLA; best-performing variant is KosMos backbone + policy head; Nature Machine Intelligence 2026

Sources · 2 sources

release_date,license,arch_notes,code_url https://github.com/Robot-VLAs/RoboVLMs Official page · Fetched 2026-09-05 · No snapshot stored yet README：initial release 12/11/24；Apache-2.0 license；KosMos VLM backbone；Nature Machine Intelligence

weights_url,license https://huggingface.co/api/models/robovlms/RoboVLMs Model hub page · Fetched 2026-09-05 · No snapshot stored yet Model card license: apache-2.0; checkpoints kosmos_ph_oxe-pretrain.pt etc.

Activity signals

Public counters fetched daily from GitHub, Hugging Face and ModelScope. Numbers only; no judgement. · fetched 2026-10-31

GitHub stars

491

+6 / 30 d

Commits · 90 d

last push 201 d ago

HF downloads · 30 d

all-time 0 · 6 likes

Weights last modified

2025-01-15

655 d ago

GitHubRobot-VLAs/RoboVLMs · 25 forks · 26 open issues · latest release v0.1.0 (2025-11-29)no push for 200 d

ModelScopeno same-name mirror found (searched by repo name)

Hugging Facerobovlms/RoboVLMs

stars · last 30 d

Public benchmark scores

As printed in the source; protocols differ between papers. Full table with notes on the benchmarks page.

Benchmark	Subset	Score	Source
CALVINKosMos P.H. (KosMos-2 backbone, policy-head, history)	abc_d	4.3 avg_len	paper · CALVIN results table (paper Extended Table 2; identical in README), row 'KosMos P.H.' train ABC test D, column Avg. Len.
CALVINKosMos P.H. (KosMos-2 backbone, policy-head, history)	abcd_d	4.5 avg_len	paper · CALVIN results table (paper Extended Table 2; identical table in GitHub README), row 'KosMos P.H.' train ABCD test D, column Avg. Len.
SimplerEnv · Google RobotKosMos P.H. (KosMos-2 backbone, policy-head, history)	move_near	0.9 fraction	paper · Extended Table 3(d), row 'RoboVLMs (Ours)', column 'Move Near / Average'
SimplerEnv · Google RobotKosMos P.H. (KosMos-2 backbone, policy-head, history)	open_close_drawer_avg	0.6 fraction	paper · Extended Table 3(d), row 'RoboVLMs (Ours)', column 'Open/Close Drawer / Average'
SimplerEnv · Google RobotKosMos P.H. (KosMos-2 backbone, policy-head, history)	open_top_drawer_and_place_apple	0.3 fraction	paper · Extended Table 3(d), row 'RoboVLMs (Ours)', column 'Open Top Drawer and Place Apple / Average'
SimplerEnv · Google RobotKosMos P.H. (KosMos-2 backbone, policy-head, history)	pick_coke_can_avg	1 fraction	paper · Extended Table 3(d), row 'RoboVLMs (Ours)', column 'Pick Coke Can / Average'
SimplerEnv · WidowX (Bridge)KosMos P.H. (KosMos-2 backbone, policy-head, history)	put_carrot_on_plate	0.4 fraction	paper · Extended Table 3(c), row 'RoboVLMs (Ours)', task 'put carrot on plate', column 'final'
SimplerEnv · WidowX (Bridge)KosMos P.H. (KosMos-2 backbone, policy-head, history)	put_eggplant_in_yellow_basket	1 fraction	paper · Extended Table 3(c), row 'RoboVLMs (Ours)', task 'put eggplant in yellow basket', column 'final'
SimplerEnv · WidowX (Bridge)KosMos P.H. (KosMos-2 backbone, policy-head, history)	put_spoon_on_towel	0.5 fraction	paper · Extended Table 3(c), row 'RoboVLMs (Ours)', task 'put spoon on towel', column 'final'
SimplerEnv · WidowX (Bridge)KosMos P.H. (KosMos-2 backbone, policy-head, history)	stack_green_block_on_yellow_block	0.2 fraction	paper · Extended Table 3(c), row 'RoboVLMs (Ours)', task 'stack green block on yellow block', column 'final'

All benchmarks →

Deployment recipe

No recipe yet. We publish only steps that actually ran on a rented GPU with a measured latency; this model has not been through that yet.

Record history

2026-09-05

Sources fetched

release_date,license,arch_notes,code_url · weights_url,license

Latency & cost

cost = p50 ms ÷ 1000 × rental ÷ 3600 × 1000

Hardware	p50 ms	p95 ms	Peak VRAM	Cost per 1k inferences ¥	samples
NVIDIA RTX 4090 ¥1.68/h · GongJi Compute (2026)	—	—	—	— no latency data	No data
NVIDIA RTX 5090 ¥2.4/h · Runyun 2.4 / GongJi 2.5 / Paratera Cloud 2.98 (2026)	—	—	—	— no latency data	No data
NVIDIA A100 80GB no rental price	—	—	—	— no latency data	No data
NVIDIA Jetson AGX Thor no rental price	—	—	—	— no latency data	No data
NVIDIA H100 80GB no rental price	—	—	—	— no latency data	No data
NVIDIA A800 80GB no rental price	—	—	—	— no latency data	No data

No latency measurements for this model yet · Data comes from literature, maintainer runs, or crowd CLI uploads. None exist so far.

Compatibility

Colored only where we hold evidence. An amber dot means the vendor declares this form as a target; that is a claim, not a verified compatibility.

	Unitree G1	Unitree H1	AgiBot Genie G1	AgileX Piper arm	SO-101	Franka FR3 / Panda	Fourier GR-2	WidowX 250 S	Google Robot（Everyday Robots）	ALOHA 2 / Cobot Magic (dual-arm)	Galaxea R1 Pro	Franka Emika Panda
RoboVLMs（KosMos P.H.）

Legend Officially supportedCommunity reproducedFine-tunable in theoryUnsupportedUnknown Vendor-declared target form 2 / 12 cells with a definite status

Reproduction records

Did this model run on your robot? Report it with a link (log, video or repo). Records are reviewed by the maintainers; a verified “works” or “works after fine-tuning” upgrades the matrix cell to “community verified” with your link as evidence. Anonymous is fine; nothing but the fields below is stored.

No verified record yet for this model.

Clear direction, grounded judgement.