Releasesby GlyphWhat's New:
We’re excited to announce a new integration between AI21 Labs and Together AI , giving developers and enterprises seamless access to Together AI’s open-source and frontier models directly within AI21 Maestro .
Also shipped Nov 27, 2025AI21 Labs in November 2025 →
feldera 0.198.0
- docs: add sidebar entry for parallel compilation docs docs: link MAX, MIN, SOME, SUM aggregates to their python test files py: add multi join tests for AS-Of Joins [adapters] Support ingesting raw data with metadata.
TiDB v8.5.4
For new features, improvements, and bug fixes released in v8.5.4 for TiDB, see TiDB v8.5.4 release notes. See the difference from the issue perspective: pingcap/tidb#62608 pingcap/tidb#59863 pingcap/tidb#62602 pingcap/tidb#61550 pingcap/tidb#63089 pingcap/tidb#57608 pingcap/tidb#63260 pingcap/tidb#58600 pingcap/tidb#63260 pingcap/tidb#62799...
More from AI21 LabsEvery AI21 Labs release →
Together AI Open-Source Models Now Available in AI21 Maestro
Release Date : November 27, 2025
Learn More:
For more details about this integration and what it enables, read the full announcement on our blog .
Export AI21 Documentation Pages as Markdown
Release Date : December 1, 2025
What's New:
Documentation pages now include a dropdown button that lets you: Copy the current page as Markdown for easy use in AI tools or other workflows. View the page as plain-text Markdown directly. This update makes it simpler for teams to reuse and share documentation content across tools and platforms.
Introducing Jamba Reasoning 3B
Release Date : October 8, 2025
What's New:
We’re excited to announce Jamba Reasoning 3B, a compact open-source reasoning model that redefines what’s possible on-device. Built on AI21 Labs’ novel SSM-Transformer hybrid architecture, this 3B-parameter model delivers 2–5× efficiency gains over competitors while achieving leading intelligence benchmarks .
Key Highlights
Smarter, faster, smaller : Hybrid SSM-Transformer efficiency enables long-context reasoning with minimal compute. Runs anywhere : Deploy locally on your phone or computer via llama.cpp or LM Studio. Enterprise-ready : Supports intelligent agentic workflows, reducing cloud costs and latency.
Get Started:
- Check out the AI21-Jamba-Reasoning-3B and Jamba-Reasoning-3B-GGUF cards. Read the full blog post.