AI.info
AI Tools Updates — Page 8
What changed in the AI tools this week — dated, from the makers’ own changelogs.
In inglese
- Ada AI Agent — Chat widget browser requirements
The chat widget now supports Chrome 99, Edge 99, Firefox 97, and Safari 15.4 as its minimum desktop browser versions.
- Ada AI Agent — URL-triggered business events removed from Embed2
Ada removed automatic URL-triggered business events from the Embed2 Web SDK.
- Vizcom — Region Maps
Segment any image into editable parts and edit color, material, and pattern region-by-region without re-rendering.
- Kilo Code — v7.7.7 (release)
Adopted OpenCode updates and added retries for transient connection drops during permission approvals.
- Keploy — v3.6.69
macOS releases are now arm64 only; replay and teardown handling were also fixed.
- Tonic Structural — v1873
Improved the cache efficiency of the Structural Agent for Anthropic models. You can now add comments to fields on Document View.
- Tonic Structural — v1874
Bug fixes and other internal updates.
- OpenHands — v1.21.0
Added support for requesting Docker execution workspaces from Docker servers.
- OpenHands — v1.22.0
Added remote MCP server testing on cloud backends, opened Git Sync to cloud org admins, and translated Turkish UI strings.
- Krisp — August Product Updates: Call Center AI
Expanded language coverage, easier Accent Conversion setup, smarter QA, and more control
- GrowthBook — v5.1.0
Added beta Contextual Bandits for personalization experiments and Saved Learnings with agent scanning and API endpoints.
- Fern SDK Generator — 3.96.2
The generated OAuth wire-test mock now expects the grant_type the generated OAuthAuthProvider actually sends.
- Fern SDK Generator — 3.96.1
Bump bundled @fern-api/generator-cli to 0.10.1; when AI analysis returns an empty changelog entry, the PR description, version bump reason, or commit message body is…
- OpenRouter — Batch API: half-price inference by bundling requests
Batch API lets you bundle requests in one POST, receive results within 24 hours, and typically pay half the per-token price.
- Lovable — Connect your app to Parallel
Search the web, read public pages and PDFs, get cited answers, and run research tasks with Parallel.
- Runway ML — Brand Kits in Runway MCP
Agents in Runway MCP can now find your Brand Kits and use those assets as references.
- n8n — n8n@2.41.0
Fixes issues including Agent Builder answer passing, workflow and credential imports, role creation, and Vault secret reads.
- Hugging Face — Transformers now runs llama.cpp quants
Transformers now runs llama.cpp quants.
- GitHub Copilot — OpenTelemetry in the GitHub Copilot app
- ChatGPT — GPT-6 Sol and GPT-6 Luna
We’re introducing GPT‑6 Sol and GPT‑6 Luna in ChatGPT Work and Codex.
- ChatGPT — Create flashcards in ChatGPT
You can now create flashcards in ChatGPT.
- Claude AI — Claude Opus 5.5 launch
We just launched Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 on most work and costs 40% less to
- Ollama — v0.34.3
GET /api/show now exposes model thinking controls and defaults; Nemotron H vision models are supported on Apple Silicon with MLX.
- Milvus — milvus-2.6.24
Milvus v2.6.24 reduces peak memory during segment loading and improves force-merge planning and filtered search.
- Milvus — milvus-3.0.2
Milvus v3.0.2 improves filtered search, group-by, and index-building performance and adds external collection and Storage V2 support.
- Amazon Bedrock — Updated the tokens-per-day quota on the bedrock-runtime endpoint to a cross-model, per-account quota
Replaced the per-model quota with a Cross-Model Max Tokens Per Day quota scoped per account, per Region across supported models.
- Amazon Bedrock — Updated the Tokens per day (TPD) definition to a cross-model, per-account quota
Updated the TPD term to describe a cross-model, per-account daily token quota across all supported Amazon Bedrock models.
- Letta — Release v0.32.15
Added a Workflow tool that wraps Agent SDK query().
- Harvey AI — Custom Knowledge Source Filters
Create a customized filter of the knowledge sources most relevant to you for reuse.
- Culture Amp AI Coach — Explore Survey Insights Without Leaving Your Report
Single-factor and single-question details now open in Insights Explorer; view trends, demographic breakdowns, and comment summaries in context.
- Jellypod — Audio Pricing: Now 50% Cheaper
Narration now costs 30 credits per minute, down from 60.
- Jellypod — Unarchive a Character
Restore an archived character to Active with their voice, scripts, and podcast assignments intact.
- Hyperscience Platform — Preserving existing field notes after upgradi...
- Hyperscience Platform — New VL M GPU image counts usage reports
Usage reports now include a # of images metric alongside the existing request count.
- Tana — From AI transcriptions to AI doing work in meetings
The AI in meetings sees what you see, remembers across every meeting, and does the work that follows.
- Culture Amp Engage — Explore Survey Insights Without Leaving Your Report
Single-factor and single-question details now open in Insights Explorer, with filters, trends, demographic breakdowns, and comment summaries.
- Scribe — Scribe desktop app 6.9.17 (Mac)
Web activity from Safari and Chrome is now supported in workflow mapping.
- Factory Droids — CLI v0.224.0, Desktop v0.181.0: Computer management in sessions, onboarding plan selection, specification downloads, and CLI and chat fixes
Added computer management in sessions, onboarding plan selection, and specification downloads.
- Slack AI — Slack 4.52.162
Bug fixes.
- Tonic Fabricate — 4.30.0
Enterprise users can now create multiple versions of a project.