AI.info
AI Tools Updates — Page 33
What changed in the AI tools this week — dated, from the makers’ own changelogs.
- Augment Code — Auggie CLI 0.36.0 Release Notes
Added a flag to control the model's reasoning effort from the CLI.
- Midjourney — Changelog - 8/20/26
You can organize projects into groups in a collapsible sidebar tree.
- Applitools Eyes — Automate document visual testing with no-code ImageTester GUI
ImageTester verifies layout and text integrity without requiring custom test code.
- Ghostwriter — Release governance: guardrails for agents at scale
Release governance brings checks, approvals, and staged rollouts from a builder's Workspace to a live customer conversation.
- Tembo — Confluence Triggers, VM Keep-Alive, and Deeper GitLab Integration
Confluence joins Jira as a trigger source; a VM keep-alive tool is available in the MCP server and CLI; GitLab automations work on merge requests opened from forks.
- LocalAI — v4.9.0
Authentication now defaults to deny; chat supports context compression, canonical model/backend pages, and video serving in vllm-cpp.
- Mistral Le Chat — API key expiration policies and notifications
Set maximum validity periods for new API keys and notify key owners before expiration.
- Instantly — 📢 Share Your Campaign Results
Share pipeline, opportunities, revenue, and positive reply rate in a card.
- GitLab Duo — GitLab 19.3 release notes
The Flow Creator foundational agent is available in the AI Catalog to create custom flows.
- PromptLayer — Deployment 1
Trace detail shows input/output previews; span details use tabs; added Gemini 3.7 Flash support.
- LiveKit Agents — @livekit/agents@1.7.0
Adds PII redaction for chat history, audio recordings, logs, and traces, plus expressive mode for voice agents.
- Sierra Insights — Release governance: guardrails for agents at scale
Release governance brings checks, approvals, and staged rollouts from a builder's Workspace to a live customer conversation.
- Suno — Offline Playlists on Mobile
Save your playlist for offline listening.
- Benchling Insights — Benchling now brings chemistry and biology together
Benchling introduces Chemistry Reactions for single-step synthesis planning and predictive ADMET models connected to the same scientific record.
- Benchling R&D Cloud — Benchling now brings chemistry and biology together
Benchling introduces Chemistry Reactions for single-step synthesis planning and predictive ADMET models connected to the same scientific record.
- Agent Studio — Release governance: guardrails for agents at scale
Release governance adds checks, approvals, and staged rollouts for changes from a builder's Workspace to a live customer conversation.
- Lokalise — See the full picture with Lokalise Analytics
Lokalise Analytics is out of beta and available in all plans, bringing Translation Quality, TM analytics, and AI Scoring together.
- NovelAI — Subscription Updates: Usage Limits & Subscription Anlas Policy Adjustments
Subscription usage limits and the Subscription Anlas policy were adjusted.
- Interprefy RSI — Interprefy Announces Speaker Labels for Clearer Multilingual Conversations
Interprefy added Speaker Labels, which identify speakers in real time and display their names in live captions and subtitles.
- Guru — Ground a skill in the exact document it should use, every time
Skill References let users attach a specific document to a Knowledge Agent skill.
- Guru — Tune how your Slack agent joins the conversation, channel by channel
Knowledge Agents in Slack now support per-channel settings for replies, thread follow-ups, and responses to people or bots.
- Mistral AI — Agentic Search. More accurate and efficient results from your AI systems.
Agentic Search provides a retrieval layer for AI systems to navigate, read, and verify information in complex documents.
- CrewAI — 1.15.17
Added declarative conversational flows and enabled declarations to drive conversational mode.
- Flux — FLUX Upscale: 2K and 4K for Video
2K and 4K for Video.
- beehiiv — AI Discovery
New playbooks for new platforms.
- CoreWeave Cloud — Update Platform
New default: 580 → 595 for GB300 NVL72 Quantum-X InfiniBand, GB300 NVL72 Spectrum-X RoCE, and RTX PRO 6000 Blackwell Server Edition.
- Julius AI — Improved Subscription Management
Users can now manage upgrades, cancellations, and scheduled downgrades through a more consistent subscription flow.
- Amazon Bedrock — Fixed Anthropic Messages API code samples and switched model-card OpenAI-SDK samples to bedrock-runtime
Fixed the Anthropic Messages API Python sample and repointed the OPENAI_BASE_URL environment variable on model cards to the bedrock-runtime /openai/v1 route.
- Grammarly — Bug fixes and performance optimization
- Raycast — 💻 Screen Awareness
AI Chat and Quick AI can use context from the focused app or window through Screen Awareness.
- RunPod Pods — Batch Jobs (BETA)
Submit inference requests to a Serverless endpoint. Create a batch, finalize it to start processing, and poll for status and progress.
- Honeycomb — Response Awareness in Automatic Investigations
Canvas Agents now receive awareness of recent alert firings during Automatic Investigations.
- Suno — Playlist Improvements on Web and Mobile
Add or reorder songs, search songs within or across playlists, pin a playlist, and view liked playlists in grid or list view.
- Tonic Fabricate — 4.25.0
You can now assign an account role (Owner, Auditor, Member) to an account API key, to manage the available actions that you can perform using the key.
- Mixedbread Rerank — Stream Agentic Search Traces
Agentic Search requests can stream live server-sent trace events when agentic search is enabled.
- Decagon Voice — Introducing Decagon Assist: Empower every representative to deliver concierge experiences
Empower every representative to deliver concierge experiences
- Recall — Recall release notes, August 19, 2026: Safari and Edge support, chat that suggests what to ask, images as cards and add 10 URLs at a time.
The Recall extension launched on Safari and Edge; OCR turns images into cards, and users can add up to 10 URLs at a time.
- Decagon AI Agent Platform — Introducing Decagon Assist: Empower every representative to deliver concierge experiences
Empower every representative to deliver concierge experiences.
- LlamaIndex — v0.14.24
Fixed issues with CodeSplitter, chat history, HTMLNodeParser, MMR search, ingestion upserts, tool parameters, filters, and AgentWorkflow output.
- LTX Studio — Meet LTX Explore: The Fastest Way to Experience LTX
LTX Explore is a self-service platform that provides instant access to LTX's video generation capabilities.