OmniThink Releases

This patch release improves stability by implementing a fallback to plain text highlighting when code language identifiers are unknown or truncated, preventing Shiki from throwing errors.

Mar 17, 2026

This minor release introduces theme customization capabilities, allowing users to specify light and dark Shiki themes via the plugin configuration options.

Mar 5, 2026

This release focuses on supporting the Qwen3.5 model family with a new Linear Attention operator, optimizing LLM resource management, and fixing memory safety vulnerabilities.

Mar 5, 2026

A major update introducing support for GPT-5.2, Gradio 6 compatibility, and robust execution mechanisms like timeouts and exception handling, alongside internationalization support.

Jan 16, 2026

This release introduces significant enhancements to code execution, sandboxing, and reliability, including support for Blaxel remote execution, custom Python executors, and exponential backoff for retries. It also adds new CLI capabilities and improves documentation with multi-language support.

Nov 17, 2025

This release focuses on maintenance updates, primarily fixing a CI bug that was incorrectly closing valid issues and updating the project readme.

Oct 28, 2025

This release focuses on enhancing executor capabilities, specifically adding support for E2B v2 and MCP structured output. It includes various fixes for model compatibility (GPT-5, Grok-4) and improvements to Docker and Modal executors.

Sep 25, 2025

Mar 9, 2025

This release removes the TensorRT integration from the project, simplifying the dependency stack.

Feb 14, 2025

The SDK introduces a new layout detection feature to improve the parsing of document structures.

Feb 14, 2025

This SDK release adds layout detection and improves text detection within auto strategies, alongside several bug fixes for configuration, testing, and heuristics.

Feb 11, 2025

Adds the layout detection feature to the core library.

Feb 11, 2025

Introduces a modular parser and formatter system, improves text detection in auto strategies, and adds type safety to strategy outputs. Also includes numerous bug fixes for configuration, testing, and logging.

Jan 16, 2025

A maintenance release focused on dependency upgrades, critical bug fixes for Streamlabs Polly and random voice generation, and improvements to the user interface.

Jun 16, 2024

Apr 22, 2024

A significant feature release introducing an autonomous agent called Data Interpreter, integrating Retrieval-Augmented Generation (RAG), and expanding support for various AI providers.

Mar 29, 2024

Minimal release notes: "A patch release for the following issues: * fixbug: #958 by @iorisa in https://github.com/geekan/MetaGPT/pull/1083"

Mar 29, 2024

Minimal release notes: "A patch release for the following issues: * feat: gemini + proxy in https://github.com/geekan/MetaGPT/pull/971 * fixbu"

Mar 12, 2024

Minimal release notes: "Version 3.2.1 is a bug release to address the following issues: - Translation - Story mode method 0 - Conversion fail"

Aug 24, 2023

This update introduces high-quality text-to-speech capabilities via ElevenLabs, adds background audio support, and enhances the user experience with one-click launch options.

May 28, 2023