AI.info
AI Tools Updates — Page 195
What changed in the AI tools this week — dated, from the makers’ own changelogs.
En inglés
- Spline AI — Tiny Updates for Spline and Hana
Change the Pivot position of 3d objects; the Code API supports material layers editing; USDZ file imports in Spline 3d editor.
- Text Generation Inference — v3.3.7
Added support for max_image_fetch_size and announced maintenance mode.
- Thematic — Product Update Q4 2025: Big Year-End Updates
Deep Dive reads every comment in a theme, identifies patterns, and writes up key insights; Answers formatting stays intact when copied.
- Waabi World — From closed course to public roads: The path to driverless
The path from closed course to public roads.
- AMD Enterprise AI Suite (SiloGen) — Version: 1.7
AIM catalogue adds gpt-oss-20b and gpt-oss-120b; Workbench supports MI325X and dynamic loading of AIMs from repo.
- SciSpace Copilot — SciSpace BioMed: AI Agent for Biomedical Research
Introducing SciSpace BioMed Agent — An AI biomedical co-scientist designed to automate biomedical research workflows.
- Poe — Added charging for canvas apps
Custom pricing for specific actions or features can now be added to canvas apps.
- Document Parse — Introducing Document Parse: Enhanced mode
Document Parse now supports three processing modes: Standard, Enhanced, and Auto.
- π0.5 — Emergence of Human to Robot Transfer in VLAs
Exploring how transfer from human videos to robotic tasks emerges in models as they scale.
- D-ID — Visual Agent Templates: Launch Faster with Pre-Built Agents
D-ID Studio now lets users create Visual Agents from templates including Brand Ambassador, SDR, Travel Guide, and Role-player.
- Framer — v3.9.0
Adds APIs for background plugin messaging, collection creation, and enhanced code file creation options.
- Poe — Added FLUX-schnell to App Creator and Script Bot Creator and improved Chinese audio generation performance
App Creator and Script Bot Creator can use FLUX-schnell for image generation and generate audio in Chinese.
- Promptfoo — Building a Security Scanner for LLM Apps
Promptfoo is adding code scanning for LLM-related vulnerabilities, first as a GitHub Action that reviews pull requests.
- Airtable Omni — Introducing Airtable for ChatGPT
- Affinda Invoice Extractor — December 2025
Added automatic routing to the correct app instance by location and released a new tables editor with grid and freeform modes.
- Affinda Resume Parser — December 2025
- Amazon Q Developer — Deprecated code transformation for GitHub
Code transformation capabilities for Java modernization have been deprecated for Amazon Q Developer for GitHub.
- Beautiful.ai — Improved AI images
Users can choose AI image generation during deck generation and select image styling from example previews.
- Reputation Reviews — Respond as a Location
This feature allows users to reply to a survey result as a location instead of as a user.
- Reputation Reviews — Support WhatsApp/Text Messaging Integration for Google Business Profile Listings
This feature allows users to configure and manage WhatsApp and SMS messaging details for their Google Business Profile listings directly within the platform.
- Reputation Reviews — Insights: AI Reputation Manager - Adding Insights Capabilities
Users can uncover deeper insights into how their brand is perceived by AI for their business questions, and track the AI response’s trend over time.
- Reputation Reviews — Integrate Upgraded Idea Portal
This feature replaces our previous feedback system with the new Idea Portal.
- Reputation Reviews — Reviews: Integrate the Reputation IQ Interface with the Reviews Page
The RepIQ button is now accessible through the updated reviews feed to open the interface directly to ask questions about reviews data.
- Reputation Reviews — Requests: Improvements
Average Star Rating in the Leaderboard report is now calculated across all reviews and surveys; follow-up messages are sent only when the survey has not yet been…
- Reputation Reviews — Campaign Notification
This feature introduces a new permission in roles.
- Poe — App Creator and Script Bot Creator now use Claude-Opus-4.5 by default
App Creator and Script Bot Creator are now powered by Claude-Opus-4.5 by default, upgrading from Claude-Sonnet-4.5.
- Sider AI Sidebar — Introducing GPT-5.2, Credit Usage Updates & Model Cleanup
GPT-5.2 is now available; Claude Sonnet 4.5 and Gemini 3 Pro are now 2 advanced credits per task; older models are being retired.
- IO Intelligence — io.net Launches the First Adaptive Economic Engine for Decentralized Compute
io.net's Incentive Dynamic Engine (IDE) is an adaptive tokenomics model bringing sustainable economics and predictable stability to decentralized GPU compute.
- Cohere — Cohere's Rerank v4.0 Model is Here!
Rerank 4.0 has pro and fast variants, supports English and non-English documents and JSON documents, and has a 32k token context window.
- IO Cloud — io.net Launches the First Adaptive Economic Engine for Decentralized Compute
io.net's Incentive Dynamic Engine (IDE) is an adaptive tokenomics model bringing sustainable economics and predictable stability to decentralized GPU compute.
- Promptfoo — GPT-5.2 Initial Trust and Safety Assessment
Promptfoo opened a PR for GPT-5.2 support.
- smart:comp 3 — smart:comp 3 is out now
smart:comp 3 launched with a Compression Matrix and a semantic approach to dynamic control.
- smart:EQ 4 — Everything That’s New in smart:comp 3
smart:comp 3 adds a Compression Matrix, a semantic approach to dynamic control, and inter-plugin communication.
- smart:EQ 4 — smart:comp 3 is out now
smart:comp 3 is out now, with a Compression Matrix and a semantic approach to dynamic control.
- Lumen5 — What Happens When You Remove All the Clicks? The Making of Chat to Video
Lumen5 created Chat to Video to reduce the decisions users face before making a video.
- Speak — Your most personalized Speak yet: What’s new in our winter release
What’s new in our winter release.
- Spline AI — Introducing 3D Shapes in Hana
Added 3D Shape support in Hana; 2D shapes stay fully editable; 3D Shape parameters can be modified between states for animation and interactivity.
- Poe — Script Bot Creator now has default images for testing
Script Bot Creator now provides default test images that can be automatically used when testing your creations.
- smart:comp 3 — Everything That’s New in smart:comp 3
Displays compressor settings; visualizes compression range; links plug-in instances; provides processing with no delay; stabilizes level variations; updates compression…
- Mobileye SuperVision — How Mobileye reimagines camera design with dual HDR image capture
Dual HDR image capture