dograh-v1.49.0
dograh-hq/dograhdograh-v1.49.0Oct 12, 2026by chewwbaka
Upgrading?
AI Summary
This release introduces Google Gemini as a new QA LLM provider, exposes configuration options for TTS speed and LLM temperature, and adds workflow editor views for model settings. It also supports multiple providers with fallbacks and improves streaming capabilities.
Key Highlights
- Added Google Gemini as a QA LLM provider
- Exposed Deepgram TTS speed and model-aware LLM temperature settings
- Added workflow editor views for model configuration and variables
- Enabled support for multiple providers and LLM fallbacks
- Implemented streaming for chat announcements and assistant replies
New Features
- Google Gemini QA LLM provider
- Deepgram TTS speed configuration
- Model-aware LLM temperature settings
- Version history in published agents
- Multiple providers and LLM fallbacks
- Streaming chat announcements and assistant replies
- Workflow editor views for model configuration
- MiniMax TTS over bidirectional WebSocket API
Full Release Notes
## 1.49.0 (2026-10-12) <!-- Release notes generated using configuration in .github/release.yml at main --> ## What's Changed ### Features * feat(qa): add Google Gemini as a QA LLM provider by @a6kme in https://github.com/dograh-hq/dograh/pull/848 * feat: expose Deepgram TTS speed configuration by @a6kme in https://github.com/dograh-hq/dograh/pull/853 * feat: expose model-aware LLM temperature settings by @a6kme in https://github.com/dograh-hq/dograh/pull/858 * feat: add version name and version history in published agents by @a6kme in https://github.com/dograh-hq/dograh/pull/862 * feat: Allow multiple providers and model configurations and enable LLM fallbacks by @a6kme in https://github.com/dograh-hq/dograh/pull/866 * feat: stream chat announcements and assistant replies by @a6kme in https://github.com/dograh-hq/dograh/pull/870 * feat: workflow editor views for model configuration, variables and settings by @a6kme in https://github.com/dograh-hq/dograh/pull/871 * feat(ui): mode picker and per-service provider tabs for model configurations by @a6kme in https://github.com/dograh-hq/dograh/pull/875 * feat: MiniMax TTS over the bidirectional WebSocket API by @a6kme in https://github.com/dograh-hq/dograh/pull/880 * feat: expand model configuration and provider voice catalogs by @a6kme in https://github.com/dograh-hq/dograh/pull/883 ### Bug Fixes * fix: return a clear 503 when the hosted service is unreachable for template workflows by @Rishxb-arch in https://github.com/dograh-hq/dograh/pull/851 * fix: finish pending tools after hangup and unify transition deferral by @a6kme in https://github.com/dograh-hq/dograh/pull/857 * fix: distinguish LLM fallback cancellations from failures by @a6kme in https://github.com/dograh-hq/dograh/pull/878 * fix: pause campaigns on provider errors and deduplicate outcomes by @a6kme in https://github.com/dograh-hq/dograh/pull/882 ### Documentation * docs: note the quickstart's public tunnel and open signup by @Rishxb-arch in https://github.com/dograh-hq/dograh/pull/850 ## New Contributors * @Rishxb-arch made their first contribution in https://github.com/dograh-hq/dograh/pull/851 **Full Changelog**: https://github.com/dograh-hq/dograh/compare/dograh-v1.48.0...dograh-v1.49.0 --- š **Deployed to Production:** https://dograh-icqa0oal9-dograh-6922fbc1.vercel.app
Related Repositories
public-apis/public-apisā
491,300
A collective list of free APIs
NousResearch/hermes-agentā
256,156
The agent that grows with you
yt-dlp/yt-dlpā
200,343
A feature-rich command-line audio/video downloader
microsoft/markitdownā
191,609
Python tool for converting files and office documents to Markdown.
Significant-Gravitas/AutoGPTā
187,707
AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.