AI.info
AI Tools Updates — Page 91
What changed in the AI tools this week — dated, from the makers’ own changelogs.
En inglés
- Fish Audio — Transcribe 1 Pro
Added Transcribe 1 Pro for short clips and long recordings, with multi-speaker transcripts and inline speaker and emotion cues.
- Suno — Albums
You can publish a body of work in an album; followers get a push notification.
- Digits AI Invoicing — Developers: Connect to the Digits API with Service Accounts
Added service accounts for connecting an organization’s systems to the Digits API without building an OAuth app.
- Slack AI — Slack 4.52.178
We cleaned the code to make sure that everything’s running smoothly.
- Tonic Fabricate — Custom MCP servers
Create custom MCP servers for mock API endpoints and sandboxes, which are temporary copies of project databases.
- Omni — September 28, 2026
Delivery emails use recipient or owner locales; users can switch attribute values, API audit events identify actors, and Agent routines can DM up to 8 people.
- Liquid Foundation Models — Introducing d1: The most capable decision model, now with vision
- AirOps — Insights citation lifecycles and Agent Analytics in connected assistants
Review citation lifecycle metrics, analyze AI crawl activity through connected assistants, and use GPT-6.1 Sol in Workflows.
- Gemini Enterprise Agent Platform — CodeMender updates (v0.12.0)
Added hidden file scanning, larger file support, report output for every scan mode, scoped fixes, and accurate session token usage.
- Tines Workflow Automation — 📊 View Tines Units over time and projected usage for current month in Monitoring
View Tines Units over time and projected usage for the current month in Monitoring
- Devin — October 2, 2026
Sidebar search now includes matching sub-sessions and workflow sessions, and the Updated-time filter no longer defaults to the last month.
- Claude Code — 2.1.290
- Devin Desktop — October 2, 2026
Sidebar search and filters were updated, and transcript displays were polished.
- Legora — Introducing Skills
Skills let legal teams encode their standards, processes and preferences inside the Legora aOS™.
- LangGraph — The LangSmith Chat model picker now offers Claude Sonnet 5.5. Sonnet 5 stays the default.
The LangSmith Chat model picker offers Claude Sonnet 5.5; Sonnet 5 stays the default.
- LangGraph — The Messages view now shows the user message on OpenAI Responses API calls that send it as a plain string, which is how calls chained with previous_response_id or a conversation send each new turn.
The Messages view shows the user message on OpenAI Responses API calls that send it as a plain string.
- LangGraph — The Messages view requests smaller pages, so long threads no longer fail with a response size error.
- LangGraph — When a trace or thread filter fails, LangSmith shows the error below the filter and keeps the selected run details open, so you can edit or clear the filter without leaving the page.
LangSmith shows filter errors below the filter and keeps selected run details open.
- LangGraph — Creating a public trace or thread share link now requires permission to read the runs, including their inputs and outputs.
Creating a public trace or thread share link requires permission to read the runs, including their inputs and outputs.
- LangGraph — Costs now resolve for every OpenAI model on the pricing page, including those listed only under “All models”, such as GPT-5.6 Luna and Terra.
Costs now resolve for every OpenAI model on the pricing page, including models listed only under “All models”.
- LangGraph — Costs now resolve for the open models Vertex AI hosts through Model-as-a-Service: DeepSeek, Qwen, GLM, Kimi, MiniMax, and GPT-OSS.
Costs now resolve for DeepSeek, Qwen, GLM, Kimi, MiniMax, and GPT-OSS hosted through Vertex AI Model-as-a-Service.
- LangGraph — The LangSmith MCP adds a query_runs tool that accepts the same parameters as the SDK’s client.runs.query() , including typed selects for costs, token breakdowns, and full inputs and outputs, with curs
query_runs accepts the SDK’s client.runs.query() parameters, with typed selects and cursor pagination; fetch_runs now returns costs and token counts.
- OpenHands Cloud — openhands: v0.77.0
Updated the runtime-api image tag to 0.11.0.
- Snowflake Data Cloud — Change the type of an organization user
- Vapi — Soniox low-confidence filter
You can now set a confidence threshold on the Soniox transcriber to discard low-confidence transcripts.
- Vapi — Chat handoff variables
Chat squad handoffs now extract variables from the chat conversation instead of overwriting valid handoff values.
- Vapi — Deepgram nova-3-pharma
Deepgram’s nova-3-pharma transcription model is now available for assistants.
- Vapi — ElevenLabs v4 Turbo
The eleven_v4_turbo voice model is now available in the dashboard voice picker and the API.
- Vapi — Visual prompt editor
The visual prompt editor no longer alters nested or multi-line Liquid tags or section headings when switching modes or saving.
- Vapi — OpenAI GPT-6 Luna
GPT-6 Luna (gpt-6-luna) is now available for assistants in the API and dashboard.
- Vapi — Simulation latency expectations
Simulations can now set latency limits for model, voice, and turn timing, fail runs that exceed them, and show measured latency results.
- Consensus — Skills
Skills run using Deep to help users search, screen, trace, and verify research.
- CodeRabbit — AI Deep Scan with linked repositories | GitHub Cloud | Essentials and above, AI Deep Scan access, Beta
AI Deep Scan can use context from linked repositories to analyze a primary repository and verify security findings.
- CodeRabbit — Repository guideline files in the review summary | PR Reviews
Review summaries list the repository guideline files used and whether each was configured, auto-discovered, or an Agent Skill.
- CodeRabbit — AI Deep Scan API access | All plans, CodeRabbit Security
The AI Deep Scan code findings API is available on all plans with a non-agentic API key and the security:read permission.
- CodeRabbit — Security merge blocking | GitHub | Advanced, Enterprise | CodeRabbit Cloud
Admins can publish a CodeRabbit Security check with a severity threshold and require it in GitHub to block merges with qualifying unresolved comments.
- Picsart AI Image Generator — Flux 3 Image in Picsart: layout control, precise edits, 4K
Flux 3 Image is now in Picsart AI Playground, with native 2K and 4K output and up to 10 reference images.
- ComfyUI — v0.39.0
Added support for LynnReal light Minimax-H3 VAE.
- LangFlow — v1.13.0.dev33
Langflow nightly v1.13.0.dev33
- Make — Deprecated metrics removed from Metricool's Get network metrics module
Metricool’s Get Network Metrics API will stop returning 30 deprecated metrics on October 5, 2026, requiring users to update affected scenarios.