Step-Audio Releases
89 releases of stepfun-ai/Step-Audio
- v2.3.1
Introduces issue reporting functionality, renames the application to 'Harper', and fixes desktop process management and font rendering.
May 30, 2026
- shadcn@4.8.3
Fixes template handling and improves Right-to-Left (RTL) CSS class transformations by preserving quote formatting.
May 29, 2026
- shadcn@4.8.2
Patch update specifically targeting improvements in how templates are processed and generated.
May 27, 2026
- 0.11.66
Adds OpenAI TTS sample rate configuration, XAI speech extensions, and fixes various ASR connection and crash issues.
May 26, 2026
- 0.11.65
Updated Cargo configuration and bumped version to 0.11.65 with minimal changes.
May 10, 2026
- 0.11.64
Added multiple TTS (Text-to-Speech) features including Rime TTS update log, Deepgram TTS extension with voice assistant integration, Tencent and Deepgram result_language_default configuration, and various TTS voice configuration fixes.
May 8, 2026
- v2.0.45
Upgraded Docker images and workflow JavaScript sandbox to Node.js 22 compatibility, fixed AI employee system prompt variable issues, and resolved multiple bugs in license settings and record import functionality.
Apr 27, 2026
- v0.9.38
This release includes multiple fixes for API key filtering, success rate tracking, array operations, DeepSeek reasoning content handling, and new features for model management and API quota monitoring.
Apr 26, 2026
- v2.0.44
Removed the recommended model from LLM service configuration and fixed a relationship field hidden sort switch issue in the client interface.
Apr 26, 2026
- v0.9.37
Added new database connection configuration with connection pool settings and introduced new channel types for Bailian Anthropic and Moonshot Coding, along with various fixes for model aliases and thinking block schemas.
Apr 24, 2026
- v0.9.36
Added Ollama channel support, improved UI rate limit display with reset times, enhanced channel probe filtering, and added Meta developer and Llama models support with various fixes for rerank and chat request handling.
Apr 23, 2026
- v0.4.12
Added curated cross-references for hub skills to improve agent routing quality, implemented Prettier markdown formatting across the codebase, and fixed documentation links and workflow command naming issues.
Apr 21, 2026
- v0.9.35
Reverted a bulk weight adjustment drag-and-drop issue, renamed model mapping terminology to model alias, forced codex channel to use array inputs, and synced model developers data.
Apr 21, 2026
- v1.8.5v1.8.5 - Spanish Locale & Token-Based Chunking
This release adds Spanish locale support and refactors embedding chunking from character-based to token-based to improve handling of CJK and mixed-language content.
Apr 19, 2026
- v0.9.34
Adds request body pass-through, improves channel health probe SQL, and accepts higher effort levels for Claude Opus 4.7.
Apr 17, 2026
- v10.4.2Nuclei Templates v10.4.2 – Release Notes
Adds 121 new templates and 61 CVEs, focusing on critical vulnerabilities like RCEs in various web applications and fixes for false positives.
Apr 15, 2026
- v2.5.17
This update adds an option to disable Claude file citations and allows specifying the GitHub repo per request, alongside fixes for workflow and HTTP/2 client issues.
Apr 15, 2026
- v2.5.16
Adds support for the LLMsTxt standard, Salesforce CRM tools, and Azure AI Foundry Claude model provider, while fixing knowledge base reading and AGUI events.
Apr 10, 2026
- v1.8.4v1.8.4 - Security Fix
A critical security release addressing vulnerabilities including Server-Side Template Injection (RCE) and arbitrary file access via path traversal.
Apr 9, 2026
- v2.5.15
Enhances workflow capabilities with skills support, nested workflows, and post-execution HITL review, along with new logging controls.
Apr 9, 2026