AI.info
AI Tools Updates — Page 191
What changed in the AI tools this week — dated, from the makers’ own changelogs.
Auf Englisch
- Replicate — Fallback model for Nano Banana Pro
Nano Banana Pro can fall back to Seedream 5.0 lite when Google’s API is at capacity.
- Harness CI — March 2026
Restricted mounting secrets through Kubernetes pod spec overlay YAML behind CI_K8S_OVERLAY_YAML_SECRET_RESTRICTION.
- Leena AI Work Assistant — Product Newsletter — February 2026
- Photoroom — The features Edit With AI and AI Beautifier are no longer in Alpha status.
Edit With AI and AI Beautifier are no longer in Alpha status.
- Empathic Voice Interface (EVI) — TTS API bug fixes
Fixed a bug where duplicate interleaved audio was included in TTS audio output.
- Photoroom — Updated
Edit With AI and AI Beautifier are no longer in Alpha status.
- Empathic Voice Interface (EVI) — EVI API additions
Added support for new supplemental LLM models and zero prompt expansion.
- Shortwave — Upgraded to Claude Sonnet & Opus 4.6
Standard and Advanced tiers now use Claude Sonnet 4.6; Expert uses Claude Opus 4.6; Basic continues to use Claude Haiku 4.5.
- qER — Qure.ai nets six new indications cleared by the FDA, taking the chest X-ray crow...
Six new indications cleared by the FDA, taking the chest X-ray crow...
- qXR — Qure.ai nets six new indications cleared by the FDA, taking the chest X-ray crow...
Six new indications cleared by the FDA.
- Photoroom — The feature Describe Any Change has been renamed to Edit With AI .
Describe Any Change has been renamed to Edit With AI.
- Photoroom — Updated
Describe Any Change has been renamed to Edit With AI.
- Smartling TMS — 🚨 Action Required: Customers must upgrade to v. 5.9.58 by February 26, 2026.
Customers must upgrade to v. 5.9.58.
- Read AI — Read AI MCP: Your Meetings Just Became Your Most Powerful Dev Tool
Read AI's MCP server is available now and you can connect in minutes.
- Read AI — AI That Works Even When You Can’t, Read AI Introduces Digital Twin
Digital Twin is introduced; cc: ada@read.ai on any email and get started for free.
- PRISMA AIRS AI Gateway — Sequential guardrails are now live!
The sequential flag runs checks in the order you define and stops the rest from executing if an earlier check fails.
- Portkey Guardrails — Sequential guardrails are now live!
With the sequential flag, checks run one after another in the order you define, and if an earlier check fails, the rest don't execute.
- π0.5 — The Physical Intelligence Layer
General-purpose physical intelligence models will enable robotics applications. Our partners are solving real-world problems.
- Nooks Dialer — Introducing AI Sequencing: the Evolution of Intelligent Outbound
AI Sequencing.
- Nooks AI Prospector — Introducing AI Sequencing: the Evolution of Intelligent Outbound
Introducing AI Sequencing.
- ClickUp Brain — Release Notes 4.02
AI Notetaker can now be sent to any call using a link.
- Trinka AI — Trinka AI adds New Checks to its Highly Used Citation Quality Report that Strengthens Research Integrity
The launch of two new checks - citations from potential predatory journals and identification of duplicate reference entries - under its Citation Quality Report.
- MLflow — MLflow 3.10.0 Highlights: Multi-Workspace Support, Multi-Turn Evaluation, and many UI Enhancements!
Support multi-workspace environments, multi-turn evaluation, trace cost tracking, Gateway usage tracking, and in-UI trace evaluation.
- Nabla Copilot — Breaking change: JWT audience us-central1 and eu-west1 are now deprecated.
JWT audiences us-central1 and eu-west1 deprecated; transcription diarization; note schemas include locale and template; normalized data coding fields changed.
- Arcee Orchestra — Trinity Large Tech Report: 400B Open MoE Architecture Explained
MoE architecture explained.
- CENTURY — API structure has changed
Added PUT /users/{userId}/external-id
- Qdrant — v1.17.0
Added relevance feedback, optimization and telemetry APIs, update queue and controls, audit logging, weighted RRF, and upsert modes.
- Saaras — Introducing Indus
- Bulbul — Introducing Indus
- Zenlytic — Introducing Web Search: real-time intelligence meets your private data
Web search is massively important for analytics agents. We built it so it’s not.
- Topaz Labs — Local Update
- PRISMA AIRS AI Gateway — 🆕 Sticky Load Balancing
Route requests with the same identifier consistently to the same target when load balancing.
- Portkey Guardrails — 🆕 Sticky Load Balancing
Route requests with the same identifier consistently to the same target when load balancing.
- Lavender — What's New in Lavender 3.0 💜
Lavender is now embedded on the left side of your compose window; Start My Email creates email options; your score is integrated into the sidebar.
- Luminar Neo — Luminar Neo 1.26.1: Improved generative AI results, stability, and new camera support
The AI Assistant panel is draggable and resizable; tools in Favorites can be reordered; adds support for Sony a7 V, Sony FX2, and Panasonic Lumix DC-S5IIX.
- PRISMA AIRS AI Gateway — Day 0 support for Claude Sonnet 4.6 ⚡
Bring Claude Sonnet 4.6 into production with AI Gateway access, guardrails, policy enforcement, observability, budgets, and rate-limits.
- Portkey Guardrails — Day 0 support for Claude Sonnet 4.6 ⚡
Bring Claude Sonnet 4.6 into production with access via the AI Gateway, guardrails and policy enforcement, observability, budgets, and rate-limits.
- HireVue Video Interviewing — Hirevue Launches Assessment Builder to Scale Validated Skills Hiring Across Enterprises
Launches Assessment Builder to Scale Validated Skills Hiring Across Enterprises
- LlamaIndex — v0.14.15
Added AgentMesh integration and support for multimodal types, prompt templates, and a multimodal chat prompt helper.
- Amazon Q Developer — Updated account limit for subscriptions
You can now subscribe users in a maximum of 20 accounts per AWS Region, per organization, increased from 10.