AI.info
AI Tools Updates — Page 10
What changed in the AI tools this week — dated, from the makers’ own changelogs.
In inglese
- Keploy — v3.6.68
- Lovable — Organize project files into collections
Group files into collections, upload files or drag them onto the tab, and see new files appear as Lovable creates them.
- Taskade — v6.253.0 — Taskade EVE Memory Controls & Voice Input
- Factory Droids — CLI v0.223.0, Desktop v0.180.0: Session forking, PyCharm, Databricks connector, streaming chat tables, and custom model credential fixes
Added session forking, Databricks setup, PyCharm, and hook risk settings; changed streaming chat tables and default autonomy.
- LiteLLM — v1.102.0: Auto Router Controls, Native OCR & Gateway Reliability
Auto router controls, native OCR, gateway reliability
- LM Studio — Bionic 1.1.5
Qwen3.8 inference using a new engine: Splash (by Inco AI); prompt-processing status for models without percentage progress.
- Claude Code — 2.1.278
Changed auto mode to default to the server-side classifier for API, Enterprise, and listed provider users; /status now shows classifier location.
- Unsloth — Docker + Multi User + AMD Support
Adds a Docker image with NVIDIA and AMD support, multi-user accounts, diffusion support, ARM64 Windows CUDA support, and RDNA1/2 support.
- Lovable — Cleaner transcripts when you dictate a prompt
When you dictate a message, Lovable removes filler words, cleans up the text as it transcribes, and recognizes product names.
- Milvus — pkg/v3.0.2
Snapshot restores targeting the same collection are now serialized.
- Milvus — client/v3.0.0
The Go client v3.0.0 adds snapshot workflows, read-only external collections, search and query features, telemetry, and structured RPC error inspection.
- Tembo — Automatic Task Recovery, Bulk Invites, and Sessions Sorting
Tasks recover and resume when a sandbox restarts; admins can invite up to 500 teammates in a batch; sessions can be sorted and filtered.
- Birdeye Reviews — Take control of your survey timelines
Set a deadline for collecting responses, close a survey, and let it automatically stop accepting responses.
- Napkin AI — v1.1.27 (2026-09-18)
Generating a visual with an OAuth access token requires the generation scope.
- Harvey AI — GPT-6 Astra Model Access
GPT-6 Astra is now available as a model provider in Harvey, giving workspaces that request it access through the model selector.
- Harvey AI — Fable 5.1 Model Access
Fable 5.1 is now available as a model provider in Harvey, giving workspaces that opt in access through the model selector.
- Meshy — Added meshy-7.1 as an ai_model value for Image to 3D , Multi-Image to 3D and Text to 3D : same parameters and credit cost as meshy-7 , higher geometry detail.
Added meshy-7.1 as an ai_model value; added geometry_resolution ( standard / 2k / 4k ).
- Birdeye Insights AI — Take control of your survey timelines
Set a deadline for collecting responses, close a survey when needed, and let it automatically stop accepting responses.
- Readwise Reader — 🔊 Fixed Feed Text-to-Speech
New items from RSS feeds that deliver plain text arrive with paragraphs, so text-to-speech works.
- Readwise Reader — 🪄 Fixed Mobile Ghostreader Prompts
Tags and highlights from Ghostreader's “Tag the document” and “Highlight the document” prompts now land on the document.
- Readwise Reader — 📥 Improved OPML Imports
The import skips feeds it couldn't fetch, keeps feed names from your file, and emails a summary listing feeds left out.
- Readwise Reader — 🧩 NEW! Reader API Highlights
You can create highlights through the Reader API; a POST to /api/v3/save/ with the document's parent_id and passage as content anchors the highlight to that document.
- Readwise Reader — 🤖 Improved Ghostreader Indexing Progress
The header says “Indexing your library...” with a running tally, then switches to the final total.
- Readwise Reader — 💌 NEW! Redesigned Readwise Emails
Daily Review and Themed Review emails have a design with highlight cards and links to editing, tagging, or sharing a highlight.
- Readwise Reader — 👆 Fixed Android Reading Taps
A fix for the invisible overlay means the page responds to taps after a jump to a chapter.
- Readwise Reader — 🧠 Improved Mobile Model Memory
A new Ghostreader chat in the Reader app starts on whichever model you used last.
- Factory Droids — CLI v0.222.0, Desktop v0.179.0: Force reboot for computers, repository panels, VSCodium, Windows MSIX packaging, and organization skill policy
Added force reboot, repository selection in panels, VSCodium, new-tab sessions, organization skill policy, and Windows MSIX packaging.
- Google AI Studio — Gemini 2.5 models access update
Limited access to 2.5 models to users who have actively used them in the past; they will continue to be served through the API.
- Decagon Voice — Introducing Generative Analytics: Turn conversation insights into a growth lever
Turn conversation insights into a growth lever
- Typefully — Typefully Is Now an Official Claude Connector
Added Typefully to Claude’s official connector directory, allowing Claude to create drafts, schedule posts, access analytics, and manage comments.
- Weaviate — v1.38.16 - Bug fixes and performance improvements across modules, roaringset, and index queues
Fixes replica movement runtime configuration and improves RoaringSet memtable flush and compaction performance.
- Fern Docs — More page suggestions in agent 404 responses
Requests for a nonexistent .md URL now return up to five “Similar pages” suggestions instead of three.
- Fern Docs — Exclude changelogs from llms.txt
Omit changelog sections from llms.txt by adding the excludeChangelog=true query parameter.
- Fern Docs — PHP filter for llms.txt and Markdown output
Pass lang=php to /llms.txt and per-page .md and .mdx URLs to return only PHP SDK code examples.
- Splice Create — Get a free cinematic synth pad for Splice INSTRUMENT
A free cinematic synth pad preset is available for Splice INSTRUMENT.
- Snyk Open Source — Announcing Snyk CLI v1.1307.3
- Replit — Connect custom APIs to Agent
Pro and Enterprise teams can use custom connectors in beta to connect Agent to APIs outside Replit’s integration library.
- Replit — Investigate with expanded audit logs
Audit logs now cover more than 65 additional events across projects, workspaces, deployments, account security, connectors, secrets, and Agent activity.
- Google Vertex AI — xAI's Grok 4.6 is generally available
Grok 4.6 is available for production use on the global endpoint and the US multi-region endpoint.
- Google Vertex AI — Agent Platform SDK for Python version 2.0.1 is available
Migrates generative AI modules to the Google Gen AI SDK, decouples the agent surface into a dedicated package, and introduces restructured namespaces.