Rerun logo

Rerun

Freemium
-(0 Отзывы)
0 Сохранено

An open-source data toolchain for multimodal robotics and physical AI data.

Веб-сайт:https://rerun.io/
Добавлено:13 июл. 2026 г.
Посетители в месяц:
83.5K
Социальные сети и электронная почта:
RoboticsPhysical AIRobot learningMultimodal dataData visualizationData querying
Посетите веб-сайт
Rerun
Посетите официальный сайт

Информация об инструменте

Что такое Rerun?

Rerun is an open-source SDK and commercial backend for logging, storing, querying, visualizing, transforming, and training on multi-rate, multimodal robotics data. It provides a shared data layer for collecting, refining, training on, deploying, and analyzing physical AI data.

Основные функции

  • Multimodal data logging and storage

    Log and store multi-rate, multimodal data as column-chunks in RRD files, either directly or by converting from other formats.

  • Interactive visualization

    Review datasets, inspect detail-level issues, and build custom views and tools with Rerun's interactive viewer and visualization framework.

  • Dataframe and SQL querying

    Query catalog metadata, columns, time ranges, and recording values using SQL or dataframes.

  • Data transformation and refinement

    Add derived columns, run computer-vision transforms, and evolve schemas without breaking historical data.

  • Training data streaming

    Define dataset mixes as queries and stream recordings directly to GPUs without an export step.

  • Open-source SDK

    Use a toolchain that includes a viewer, query library, visualization framework, CLI, file format, and APIs for Python, Rust, and C++.

  • Rerun Hub backend

    Use the commercial Rerun Hub backend for data cataloging, byte-range indexing, object-storage retrieval, metadata management, and streaming at scale.

  • Shared team data access

    Rerun Hub supports team access with SSO and authenticated share links for shared viewing, annotation, and failure tracing.

Варианты использования

  • #1

    Robot learning data workflows

    Collect, refine, train on, and deploy multi-rate, multimodal robotics data through one data layer.

  • #2

    Dataset review and debugging

    Review datasets, inspect detail-level issues, visualize recordings, and trace deployment failures back to the data that caused them.

  • #3

    Robotics data querying

    Query recordings with SQL or dataframes across catalog metadata, columns, time ranges, and recording values.

  • #4

    Direct model training from recordings

    Define dataset mixes as queries and stream recordings directly to GPUs without an export job.

  • #5

    Physical AI application development

    Build applications and compute workflows for physical intelligence using the open-source Rerun SDK.

Лучше всего для:Physical AI teamsRobot learning teamsRobotics engineersAI researchersDevelopers working with multimodal robotics data

Тарифные планы

The Rerun SDK is open source. Rerun Hub is identified by the official source as commercial.

Альтернативы

Google AI Studio

Google AI Studio

Browser-based workspace for developers to prototype with Gemini models, test prompts and build AI applications.

🇺🇸12.1%
117M
Инструменты для разработчиковGoogle AI Studio+1
MarsCode

MarsCode

Cloud‑based IDE with a built‑in AI assistant for code completion, generation, debugging and one‑click deployment.

🇨🇳21.3%
79K
Инструменты для разработчиковAI coding assistant+8
Replit

Replit

AI app-building platform for founders and developers to create, edit and publish applications from natural-language instructions.

🇺🇸29.7%
13M
Инструменты для разработчиковReplit+1
Cursor

Cursor

AI coding agent for context-aware software development

🇺🇸19.9%
22M
Инструменты для разработчиковAI coding assistant+10
RunPod

RunPod

GPU cloud infrastructure for training, inference, and scalable AI workloads.

🇺🇸28.4%
2.2M
Инструменты для разработчиковAI developer cloud+11
Traceable

Traceable

Intelligent Application and API Security at Enterprise Scale

🇺🇸23.4%
31K
Инструменты для разработчиковAPI Security+11
FlexAI

FlexAI

Agent-native AI infrastructure for managed inference on open models via one OpenAI-compatible key, with Agent SDK and private cloud options.

🇺🇸30.8%
18K
Инструменты для разработчиковAI infrastructure+11
Retool

Retool

Build and manage production-ready internal applications with connected data and AI capabilities.

🇺🇸29.3%
1.6M
Инструменты для разработчиковInternal tools+11

Часто задаваемые вопросы

What is Rerun?
Rerun is an open-source data toolchain and commercial backend for working with multi-rate, multimodal robotics and physical AI data.
What can Rerun be used for?
Rerun can be used to collect, store, query, visualize, transform, refine, and train on robotics data, as well as review datasets and investigate deployment failures.
Is Rerun open source?
The Rerun SDK is open source. Rerun Hub is identified by the official source as a commercial backend.
Which programming languages does the Rerun SDK support?
The official source describes APIs for Python, Rust, and C++.
Can Rerun stream training data directly to GPUs?
Yes. Rerun supports defining dataset mixes as queries and streaming recordings directly to GPUs without an export step.
Where can Rerun be used?
The official source lists local, self-hosted, and cloud options.
Does Rerun support SQL queries?
Yes. Rerun supports SQL and dataframe queries across catalog metadata, columns, time ranges, and recording values.

Rerun Контакты и социальные сети

Посетите официальный сайт

Отзывы

-

Поделитесь своим опытом использования этого инструмента...

Информация об инструменте

Модель оплатыFreemium
КатегорияИнструменты для разработчиков
Теги
RoboticsPhysical AIRobot learningMultimodal dataData visualizationData querying
Языканглийский
Последнее обновление16 июл. 2026 г.
Заявить права на этот инструмент