# RoboWild Benchmark: Claude Fable 5.2 Leads at 13.3% While RoboDojo Hits 64.9%

> Source: <https://explainx.ai/blog/robowild-fable-5-2-13-percent-xlang-robot-benchmark-november-2026?utm_source=rss&utm_medium=feed&utm_campaign=blog>
> Published: 2026-11-21 00:00:00+00:00

**Recommendation:** Evaluate robot-use agents on dense real-world-style scenes like RoboWild's 341 tasks rather than simulation alone; the same model family dropped from 64.9% on RoboDojo Sim to 13.3% on RoboWild.
More in the playbook "Evals and testing for LLM apps": <https://wpnews.pro/playbooks/evals-testing>

**Claude Fable 5.2 tops the public [RoboWild](https://robowild.xlang.ai/?ref=explainx) leaderboard at 13.3% weighted success**, yet that is still only about **one task in eight** across **341** real-world-style robot chores. The same model family scores **64.9% on RoboDojo Sim** in [Anthropic's November 9 launch table](https://explainx.ai/blog/claude-fable-5-2-mythos-5-2-launch-pricing-api-2026). **RoboWild is the gap made visible**: dense rooms, articulated objects, and safety side effects that simulation stacks smoothed away.

The benchmark comes from **XLANG Lab**, the group behind **[OSWorld-style computer-use stacks](https://explainx.ai/blog/alibaba-opengcua-open-stack-generalist-computer-use-agents-osworld-v2-2026)**. Two years after OSWorld, frontier models handle many **digital** workflows; **physical robot-use agents (RUAs)** remain the **next frontier**, in the lab's wording on the site.

## TL;DR: scores and what they mean

| Question | Answer | 
|---|---|
| Best model on RoboWild? | Claude Fable 5.2 at **13.3%** (ReAct, highest thinking) | 
| Runner-up? | GPT-6 Astra **11.6%** ; Gemini 4 Carbon**10.6%** | 
| Tasks / scenes? | **341** tasks,**38** furnished scenes,**4,480** objects | 
| Robots? | Mobile ALOHA, Franka Mobile, Unitree G1, Dexmate Vega-1 | 
| Top failure mode? | **Grasp failure** 42.6% of failed runs | 
| Sim comparison? | Opus 5.5 **50%** RoboDojo vs**4.81%** RoboWild (100 calls) | 
| Assets? | [Hugging Face dataset](https://huggingface.co/datasets/xlangai/RoboWild?ref=explainx) (~18.5 GB, Apache-2.0) | 

## What RoboWild adds beyond RoboDojo and BEHAVIOR

RoboWild home entryway furnished scene for robot-use tasks
*Image: [XLANG Lab](https://robowild.xlang.ai/?ref=explainx)*

[RoboWild's site](https://robowild.xlang.ai/?ref=explainx) argues existing robotics benchmarks were **not demanding enough**: small workspaces, simple object sets, and **narrow task distributions**. RoboWild targets **wild** robot use: **homes, workshops, laboratories, and stores** with **3.60 objects per square meter**, which the authors report is **nearly twice** the next densest benchmark in their comparison table (**Gym-Anything** at 1.82).

Coverage numbers from the official page:

- **23 scene types** versus single-scene or low-diversity suites.
- **61.9%** of tasks need**at least two** of eight measured skill types.
- **65.7%** require**dependency chains** of two or more subgoals.
- **58.6%** need**visual disambiguation** among similar objects.

Task examples on the demo page include mundane but hard chores: **stack mail in a drawer**, **hang umbrellas on specific hooks**, **fold glasses onto a plate**, **open a locked door**. Those are the tasks humans dismiss as "easy" until a robot must do them with **three wrist cameras** and a **JSON tool interface**.

If you tracked **[Waddle-1 on RoboDojo-RC](https://explainx.ai/blog/waddle-1-robodojo-rc-81-score-claimed-vs-verified-november-2026)** this week, read RoboWild as the **opposite stress test**: not **six real bimanual tasks** on a leaderboard row, but **hundreds of diverse scenes** where **87% of runs disturb unrelated objects**, per the site's safety statistics.

## Leaderboard: Fable 5.2 first, everyone still low

The published **ReAct** leaderboard on robowild.xlang.ai (five models, highest thinking per model) ranks:

| Rank | Model | Weighted success | 
|---|---|---|
| 1 | Claude Fable 5.2 | **13.3%** | 
| 2 | GPT-6 Astra | 11.6% | 
| 3 | Gemini 4 Carbon | 10.6% | 
| 4 | Claude Opus 5.5 | 6.5% | 
| 5 | GPT-6.1 Sol | 4.6% | 

The abstract on the same page notes **token economics**: **Fable 5.2, Opus 5.5, and Carbon** consume **several times more output tokens** than **GPT-6.1 Sol and GPT-6 Astra** per task. Routing logic that picks **cheap models** for coding may need a **different policy** for **multi-minute robot episodes**.

Anthropic's launch post already showed **Fable 5.2 at 64.9% on RoboDojo Sim** versus **39.4% for Opus 5.5** and **27.6% for Fable 5.1**. RoboWild **preserves the ordering** (Fable above Opus) but **crushes absolute success**: the site plots **Opus 5.5 at 50.00% on RoboDojo and 4.81% on RoboWild** with **100 tool calls**.

That is the headline for **[top Fable 5.2 demos](https://explainx.ai/blog/top-20-things-built-with-claude-fable-5-2-matt-shumer-anthropic-mollick-2026)** too: impressive **puzzle cubes and bottle caps** in curated clips are **not** the same distribution as **341 wild tasks** with **grasp failures** dominating.

## Task design: perceptual and physical skill mix

RoboWild publishes a skill requirement breakdown on robowild.xlang.ai that explains why **13% success** is not a formatting error.

**Perceptual demands** (share of tasks requiring the skill):

- **Visual disambiguation** : 58.6% — picking the right mug, hook, or letter among lookalikes.
- **Visual reasoning** : 14.6% — layouts and spatial relations beyond OCR.
- **Text reading (OCR)** : 13.4% — addresses on mail, labels on controls.
- **Memory** : 5.0% — carrying state across subgoals.

**Physical demands**:

- **Articulation or material interaction** : 53.6% — doors, dials, deformable items.
- **Precise manipulation** : 48.5% — stacking, folding, placement tolerances.
- **Navigation** : 28.5% — moving bases through clutter.
- **Bimanual manipulation** : 21.3% — two-arm coordination.
- **Physical tool use** : 12.6% — drills, stand mixers, vise-like objects listed in the interactive object gallery.

**61.9%** of tasks require **at least two** skill types; **0.8%** require five or more. That composition is why **computer-use agents** trained on **flat screenshots** do not transfer cleanly: the observation stream is **multi-camera**, **temporally extended**, and **physically consequential**.

## Harness: how a VLM actually drives the robot

RoboWild documents a **ReAct-style** loop on robowild.xlang.ai:

1. The model receives a **task instruction** ,**robot-use guidance** (frames, units, quaternions),**tool schema** , and**three images** (head, left wrist, right wrist).
2. It calls **`move_robot`** (and**` done`** ) — one tool per turn.
3. The simulator advances and returns **observation images** until termination.

Embodiments share that interface across **mobile dual-arm**, **Panda on a base**, **humanoid**, and **wheeled humanoid** platforms. The benchmark is explicitly about **cross-embodiment generalization**, not a single arm in an empty bench.

Ablations mentioned on the site (not fully tabulated on the landing page) include **adding depth** and **enabling code execution** as ways to **raise success**. Builders replicating OSWorld-style **computer-use** stacks should assume **robot-use** needs **similar harness work** plus **physics-side safety**.

## Failure modes: grasp, side effects, and humanoid pain

RoboWild's results section is unusually blunt for a leaderboard launch:

- **Grasp failure** :**42.6%** of the**188 failed runs** analyzed.
- **Motion or execution failure** :**28.7%** .
- **Functional object operation failure** :**16.0%** .
- **Goal or target selection error** :**6.9%** .

**Failure recovery** rates differ by error class: **Interface** errors recover locally **78.7%** of the time; **Contact** errors only **25.7%**. Agents can **retry UI-like mistakes** more often than **physical slips**.

The **safety and side effects** block is what product teams should quote:

- **Unrelated objects disturbed** :**87%** of runs.
- **Collision stops** :**78%** .
- **Task objects dropped or tipped** :**75%** .

That aligns with regulatory attention on **robotaxi perception** after **[California's Waymo freeway order](https://explainx.ai/blog/waymo-california-dmv-freeway-ban-november-2026)**: low success is not only **task failure** but **collateral motion** in cluttered spaces.

The site also states VLMs **struggle on the humanoid embodiment** specifically, while **depth** and **code execution** help in separate harness tests. If you evaluate **[Perceptron Isaac checkpoints](https://explainx.ai/blog/perceptron-isaac-0-5-ready-to-run-checkpoints-yam-lerobot-november-2026)** for **VLA policies**, RoboWild is the complementary question: **what happens when a general VLM steers the same hardware through a tool API?**

RoboWild embodied AI evaluation benchmark comparing sim scores to wild robot tasks

## How to try RoboWild assets and code

Primary sources:

- **Site and leaderboard** :[robowild.xlang.ai](https://robowild.xlang.ai/?ref=explainx)
- **Paper PDF** : linked from the site citation to Hugging Face (`RoboWild.pdf` on the dataset repo)
- **GitHub** :[xlang-ai/RoboWild](https://github.com/xlang-ai/RoboWild?ref=explainx) (installation guide referenced from HF)
- **Dataset** :[xlangai/RoboWild](https://huggingface.co/datasets/xlangai/RoboWild?ref=explainx) —**18.5 GB** OpenUSD scenes, objects, robots, materials for**Isaac Lab**

Quick start from the dataset card:

```
git lfs install
git clone https://huggingface.co/datasets/xlangai/RoboWild robowild_assets
export ASSETS_DIR="$PWD/robowild_assets"
```

Then follow the **RoboWild installation guide** on GitHub for simulator setup. The repo layout lists **scene01–scene38**, reusable **objects/**, **robots/** for all four embodiments, and shared **materials/**.

There is **no substantive Hacker News thread** yet under the RoboWild name (search checked November 21, 2026). For developer reactions on **robot evals**, see **[uni-robot-agent harness effects](https://explainx.ai/blog/uni-robot-agent-open-qwen-robodojo)**.

## Benchmark landscape: density and scene breadth

RoboWild's comparison table on the official site positions it against **RoboCasa365**, **BEHAVIOR-1K**, **Embodied First Exam**, **LIBERO**, **RoboARCade**, and **RoboDojo** on **scene types** and **objects per square meter**. RoboWild reports **23 scene types** at **3.60 objects/m²**; **RoboCasa365** shows **1 scene type** at **1.41 objects/m²** in the same table.

That is not an attack on prior benchmarks — each suite serves a purpose. **RoboDojo** remains the **simulation slice** Anthropic uses in launch materials and **RoboCurve** adapts for **real arms**. RoboWild is the **stress test for generalist VLMs** that must **pick tools, disambiguate clutter, and avoid side effects** in furnished spaces.

Teams comparing **[AA-Robotics humanoid scores](https://explainx.ai/blog/aa-robotics-v0-1-frontier-models-robot-arm-humanoid-benchmark-claimed-vs-verified-2026)** or **[Gemini Robotics factory demos](https://explainx.ai/blog/gemini-robotics-2-5-mercedes-kitting-asimov-jailbreak-2026)** should add RoboWild when the question is **"can a frontier chat model operate my robot API?"** rather than **"does a specialized VLA policy track boxes?"**

## What to do with these numbers in production

Practical takeaways if you ship **robot-use** or **VLA** features:

1. **Treat sub-15% success** as the**current ceiling** for general VLMs on**wild** tasks, not an outlier.
2. **Budget tokens and wall clock** per episode; Fable-class models**spend more output tokens** per task on the official abstract.
3. **Log side effects** , not only task success —**87% disturbance rate** is a**liability** metric.
4. **Compare harnesses** , not raw model names: RoboDojo, RoboDojo-RC, and RoboWild use**different scenes, caps, and rubrics** .
5. **Pair sim training** with**furnished-scene eval** before marketing "general robot intelligence."

For **[Fable versus Opus routing](https://explainx.ai/blog/claude-fable-5-2-vs-opus-5-5-benchmarks-cost-routing-which-model-2026)**, RoboWild reinforces **Fable for long-horizon physical tool loops** — but **only if** your product tolerates **cost per attempt** and **low absolute success** today.

Frontier comparisons that include **Gemini 4 Neon** and **GPT-6 Astra** live in our **[November frontier table](https://explainx.ai/blog/gemini-4-neon-vs-claude-opus-5-5-vs-gpt-6-astra-vs-carbon-comparison-november-2026)**; RoboWild adds **robot-use** as a row you should **not** infer from **coding benchmarks** alone.

## Related on explainx.ai

*Leaderboard figures and failure statistics are accurate as of November 21, 2026 per [robowild.xlang.ai](https://robowild.xlang.ai/?ref=explainx) and [Anthropic's Fable 5.2 announcement](https://www.anthropic.com/claude-fable-and-mythos-5-2?ref=explainx).*
