Releasesby GlyphMeasuring how AI agents handle everyday wet-lab science
OpenAI released BioBasicsBench, a 100-task benchmark designed to evaluate how well AI agents interpret raw experimental data and perform tasks related to wet-lab science.
Friday, November 6, 202626 releases
Kimi K3.1 Pro (Highspeed) now available on Amazon Bedrock
Moonshot AI's Kimi K3.1 Pro is now available on Amazon Bedrock. The model features a 1-million-token context window and supports prompt caching for agentic tasks like coding and document processing.
Devin MCP: Run Automations and Oncall as a Service User
Devin can now execute automations and oncall responders as specific service users, reducing configuration drift in Terraform.
Chat SDK agents can now stream generative UI with json-render
Vercel introduced json-render support in Chat SDK, enabling AI agents to stream structured JSON that renders into platform-native UI cards on Slack and Teams.
What's new in Sigma
Sigma enabled AI Agents to set temporary control values during automated or API runs, and updated the Sigma MCP server to allow assistants to override query filters.
Dedicated Memory Bank functionality in the Google Cloud console
Users can now create, configure, and inspect Agent Platform Memory Bank instances through a dedicated page in the Google Cloud console. The update provides centralized memory management, including structured and unstructured memory filtering and revision history.
Thursday, November 5, 202611 releases
Interactions API long decodes for Gemini 4 Argon
Google Cloud announced general availability for long decodes up to 1,048,576 tokens on the Interactions API for Gemini 4 Argon models using automatic continuation or background execution.
Announcing Web Answer on Amazon Bedrock AgentCore
Web Answer is now generally available on Amazon Bedrock AgentCore. It provides a managed tool that returns synthesized, cited answers grounded in web sources in a single call.
Amazon OpenSearch Service optimized engine now supports agent traces
The Amazon OpenSearch Service optimized engine now supports agent and OpenTelemetry traces. This allows users to correlate logs and traces in a single view for debugging agent sessions.
External knowledge sources are in Beta
Databricks introduced beta support for external knowledge sources, allowing users to pull metadata from Power BI semantic models into Databricks. Genie can now use these definitions to answer questions.
Use GPT-6-Cyber to investigate vulnerabilities, reproduce them in an authorized test environment, and verify
OpenAI has released GPT-6-Cyber, a model designed to investigate vulnerabilities, reproduce them in authorized environments, and verify fixes.
Dynatrace Release Radar 10.26
Dynatrace released updates including an SRE Agent for automated problem investigation and evidence gathering. The update also adds new workflow templates for alert reduction, Kubernetes troubleshooting, and database operations.
Le Chonk brings the brains. Studio puts it to work
Mistral AI introduced Mistral Studio and Le Chonk to enable building and deploying enterprise agents, workflows, and applications with control over models, data, and infrastructure.
November 2026
Perplexity added an ultra effort level option to its Agent API, designed for broad research and data-gathering tasks using Claude Opus 5.5, with support for background execution mode.
Improvements to agent-generated vector illustrations and icons
The Figma agent now generates editable, style-consistent vector illustrations and icons.
Streamlit 1.67.0
Streamlit 1.67.0 released features including an HTTP Agent API, cache budgets with server.maxCacheDataSize, Arrow serialization for ECharts, and input widget enhancements.
arize-phoenix-client: v3.7.0
Arize Phoenix client v3.7.0 adds model parameter and agent setting recording to experiment metadata and includes validation for annotation dataframes.
Wednesday, November 4, 202615 releases
Introducing Images 2.6
Images 2.6 is now available, offering faster image generation, GIF creation, and improved editing capabilities across ChatGPT, Codex, and the API.
AlloyDB Express is available in Preview for PostgreSQL 18
AlloyDB Express is available in Preview for PostgreSQL 18. Optimized for rapid prototyping and for workflows that use AI coding agents, this configuration provides the following features: Fast provisioning Scale-to-zero compute when idle A monthly free tier for the first 12 months of usage One-click upgrades to AlloyDB for PostgreSQL For more information...
Gemini CLI Release v0.67.0
- Changelog for v0.65.0 chore(release): bump version to 0.67.0-nightly.20261020.gb041de017 Changelog for v0.66.0-preview.0 feat(cli): bridge the IDE companion into container sandboxes over exec stdio fix(core): prevent duplicate agent loading when workspace is home directory fix(ide): exclude IDE settings and GEMINI.md files from open file con text...
Semantic Search in Airbyte Context Layer
Semantic search is now available for Airbyte sources synced to BigQuery and Snowflake. This feature allows AI agents to query data based on meaning rather than keywords.
Moonshot AI Kimi K3.1 Pro is now available on Unity Gateway
Databricks added Moonshot AI Kimi K3.1 Pro to Unity Gateway on pay-per-token endpoints in us-east-1, offering native vision and reasoning capabilities for agent workflows.
AI Function tasks for jobs (Beta)
Databricks jobs now support AI Function tasks for parsing, extracting, or classifying files and table rows without writing SQL, currently in beta.
MLflow 3.18.0
MLflow 3.18.0 helps you reuse agent expertise, find regressions before shipping, and monitor quality in production. Major New Features Automatic Issue Detection: Turn recurring agent failures into issues you can investigate and fix.
Compare AI agent changes with Online Experiments in Agent Observability
Datadog Online Experiments allows engineers to test AI agent changes against production traffic. It enables comparison of response quality, cost, latency, and error rates.
Zed v1.27.0
This week's release includes syntax highlighting for Go 1.27 generic methods, an option to open README files by default, and AI commit messages that follow your repository's recent commit style.
Dify 1.19.0 - Trace Your Workflows, Layer by Layer
Dify 1.19.0 brings nested workflow tracing to your observability tools. Follow a workflow into the Workflow Tools it calls, inspect the steps inside each invocation, and keep the LLM context alongside the execution tree.
Background runs and additional supplementary tools for agents in the Snowflake-managed MCP server
Snowflake added a background boolean parameter to CORTEX_AGENT_RUN and CODING_AGENT tools in its managed MCP server to enable non-blocking, pollable agent executions.
Kimi K3.1 Pro now has more ZDR providers on AI Gateway
Vercel AI Gateway expanded Zero Data Retention provider support for Moonshot AI Kimi K3.1 Pro and K3.1 Pro Highspeed models, supporting 1M token contexts and multimodal input.
Dreaming: Lovable keeps its memory up to date
Lovable now automatically tidies project memory and AGENTS.md files in the background for paid plans, ensuring project rules remain current.
e2b@2.63.0
Version 2.63.0 of e2b adds a strict mode to the pause function, preventing sandbox pauses if the guest agent is unresponsive.
Tuesday, November 3, 202621 releases
DeepSeek V4.1 Flash (1017) support in AIP through Fireworks
AIP added experimental support for DeepSeek V4.1 Flash (1017) through Fireworks on non-georestricted commercial enrollments. The open-weight multimodal model supports text and image inputs, tool calling, and a 1-million-token context window with US-hosted inference.
Governance workflows in Knowledge Catalog support batch metadata change requests
Governance workflows in Knowledge Catalog support batch metadata change requests. You can group up to 250 create, update, and delete operations for entries and entry links into a single change request and curate batch proposals generated by the enrichment agent before approval. This feature is available in preview .
Tencent Hy4 now available on AI Gateway
Hy4 from Tencent is now available on AI Gateway , at 50% off through December 31, 2026. Hy4 supports reasoning , tool calling , structured outputs , and image input, and has a 1M-token context window for work across a codebase or many documents. Hy4 is the first Hy model with vision. It can pick out fine detail in an image and read complex charts.
Vercel Security Review now in public beta
Vercel Security Review is now in public beta for Pro and Enterprise teams. Once enabled, every push to a connected GitHub repository is automatically reviewed for vulnerabilities, and findings are reported in the Vulnerabilities dashboard . What triggers a review Security Review runs automatically when code is pushed to a connected GitHub repository.
Inference New Serverless mode: Ultra-fast (Preview) Ultra-fast is a new high-speed Serverless mode, now in...
Inference New Serverless mode: Ultra-fast (Preview) Ultra-fast is a new high-speed Serverless mode, now in Preview, for highly latency-sensitive applications. Ultra-fast variants aim for 6-8x generation speed, and up to 10x for agentic coding tasks. Like Fast, it is not a different model, so output quality is the same.
App testing
Live on cloud • All plans Public beta You can now use the Build with tests setting to validate that your app is functioning as expected and catch unintended changes before you publish your app to end users. Toggle this setting on from the chat settings or from the debug console . note App testing is currently rolling out to cloud organizations.
OneLake: The file system for agent-first data engineering
OneLake now offers a remotely hosted MCP server in preview, allowing AI agents to manage data and shortcuts without a local server installation.
Dependabot version updates support Agent Package Manager
You can now keep your AI agent skills, prompts, and instructions up to date with Dependabot version updates for Agent Package Manager (APM) . Dependabot opens pull requests to update supported git dependencies declared in apm.yml and regenerates apm.lock.yaml when it’s present. This support covers version updates, not security updates.
GitHub Copilot is now available in JetBrains Air
You can now delegate coding tasks, guide execution, and review results with GitHub Copilot in JetBrains Air . The integration uses the Agent Client Protocol (ACP) and supports tasks in your JetBrains IDE or in the cloud. Use Copilot in Air alongside the GitHub Copilot plugin and choose the workflow that works best for you.
UI improvements now generally available for PD Reliability Platform customers
Customers on PD Reliability Platform now have an improved UI experience, including: Consolidated settings: Brings all your configuration and admin settings together in one place. New nav: Provides streamlined access to your most important capabilities. Your Subscription: Shows where the opportunities are to shift work off humans and onto AI and automation.
Agents SDK now supports Pi 1.5 with fewer storage writes and lower memory use
Agents SDK v0.33.0 upgrades PiHarness to Pi 1.5 ↗︎ . The release reduces storage transactions while streaming, improves throughput across many sessions, and reduces memory use when streaming large tool calls. Performance improvements Far fewer storage writes while streaming.
Lovable follows AGENTS.md files in your project's subfolders
Lovable now follows the AGENTS.md and CLAUDE.md files in your project's subfolders . When Lovable opens or edits a file, it reads the instruction files in that file's folder and in every folder above it.
Your prototype connections, now context for code layers
When you create a code layer from design frames, the agent now reads the prototype connections and interactions between them.
MCP server: Send feedback and questions to PlanetScale
We've added two tools to the PlanetScale MCP server that let your AI agent tell us what got in its way. planetscale_submit_feedback : report a problem, such as a wrong result, a confusing error, a documentation gap, or a missing tool. planetscale_ask_question : record a question the docs and tools couldn't answer, so we know where to fill the gap.
Your agent can now tell Netlify when it gets stuck
Your agent can now tell us when something gets in its way. All you need is to have the Netlify MCP Server installed or set up. If you're using a Netlify Connector or plugin with your agent, you don't need extra setup here.
Evaluate your AI agents with Fireworks AI models in Datadog Agent Observability
Users can now evaluate AI agents and compare model performance within the Datadog Agent Observability Playground using integrated Fireworks AI models.
- Redesigned MotherDuck UI: The MotherDuck UI has a new look and feel
- Redesigned MotherDuck UI: The MotherDuck UI has a new look and feel. Home opens with an Agent Chat prompt and a setup tile, while the updated sidebar groups Recents , Favorites , and Attached Databases . Every page also has Support chat in the top right to enable users to get help from our Support team.
Monday, November 2, 202616 releases
CoCo automations (General availability)
CoCo automations are now generally available and are no longer in Preview . Automations let you schedule recurring, unattended CoCo runs that execute in a Snowflake-managed sandbox. Each run creates a Cortex thread that you can open to inspect the agent's messages, tool calls, and results.
Knowledge Studio in Knowledge Catalog is available in Preview
Knowledge Studio in Knowledge Catalog is available in Preview . Knowledge Studio lets you explore data assets and semantic relationships using natural language and an interactive context graph, run automated enrichment tasks to generate and curate metadata from schemas, query logs, and enterprise knowledge signals, and package curated domain knowledge into...
The Data Engineering Agent now supports BigQuery AI, ML, and object functions and statements in generated...
The Data Engineering Agent now supports BigQuery AI, ML, and object functions and statements in generated Dataform SQLX pipelines and agent instruction files . This feature is generally available (GA).
From a prompt to an operational dashboard: Generate Real-Time Dashboards with Copilot (Preview)
Copilot in Microsoft Fabric can now generate Real-Time Dashboards from natural-language descriptions. It uses database context from an Eventhouse to create initial dashboard layouts.
Long conversations and agent loops on the Messages API can now send smaller requests and get a lower time to
Long conversations and agent loops on the Messages API can now send smaller requests and get a lower time to first token with message threads, in beta with the message-threads-2026-08-12 beta header. The API keeps the conversation on the server, so each later request sends only the new messages.
Datadog is now available on Anthropic’s Claude Marketplace
Datadog is now available on the Claude Marketplace, providing visibility into Claude-powered AI applications.
LongCat 2.5 now available on AI Gateway
LongCat 2.5 from Meituan is now available on AI Gateway. This model combines native visual understanding with coding and multi-step tool use for agents working across complex codebases and full-project builds.
All changes: ElevenLabs CLI
- v1.9.0 - agents workflow diagram draws an agent workflow in the terminal, or prints it as Mermaid with --format mermaid or --format mermaid-md. agents graph shows how agents, tools, tests, knowledge base documents and handoffs connect across the agents in agents.json, and --uses lists the agents that reference a given tool or document.
Trace propagation across triggered builds
Agent spans across your multi-pipeline workflows can now share a single OpenTelemetry trace without contacting Support to enable propagation. Buildkite automatically passes trace context from a parent build to builds it triggers. This also works for internal trigger chains without externally supplied trace context.
Browser applications now show Request errors
Find the failed API calls behind broken pages with Browser Request errors In your Browser application, a new 'Request errors' page groups AJAX and fetch requests that returned an HTTP 4xx or 5xx status code. You can see which backend or third-party calls are failing, how often, and where to start debugging. 'Errors Inbox' is also now 'JavaScript errors'.
Sunday, November 1, 20261 release
Atlas Agent Engine in public preview
Atlas Agent Engine is now available in public preview. It provides a secure way to run AI agents at scale.
Friday, October 30, 20261 release
Ling 3.1 Flash 1019 now available on AI Gateway for free
InclusionAI's Ling 3.1 Flash 1019 model is now available on AI Gateway, featuring a 1M-token context window for coding and agent workflows.