{"id":312,"date":"2025-04-29T08:30:38","date_gmt":"2025-04-29T15:30:38","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/foundry\/?p=312"},"modified":"2025-04-29T13:53:48","modified_gmt":"2025-04-29T20:53:48","slug":"whats-new-in-azure-ai-foundry-april-2025","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/foundry\/whats-new-in-azure-ai-foundry-april-2025\/","title":{"rendered":"What’s new in Azure AI Foundry | April 2025"},"content":{"rendered":"

TL;DR<\/h2>\n

Long-context GPT-4.1, GPT-image-1, new o-series reasoning and GPT-4o audio models headline this month\u2019s releases. On the agent side we get cross-cloud A2A, BYO thread storage, an MCP server starter, and a turnkey AI Red Team. Developers also gain a VS Code extension, richer evaluation metrics, persistent memory via Mem0, a full RAG demo suite, and new Content Understanding & Document Intelligence endpoints\u2014everything you need to build, test, and ship safer GenAI apps on a single platform.<\/p>\n

Join the new Azure AI Foundry Developer Forum on GitHub<\/h2>\n

We launched the new GitHub Discussions Developer Forum last week and we’re inviting you to connect with engineers and peers to ask questions, showcase your projects, vote in polls, and shape the roadmap\u2014all in one place. Bring your ideas, code, and curiosity!<\/p>\n

Open Discussions<\/a><\/div><\/p>\n

Models<\/h2>\n

GPT-4.1 One-Million-Token Context<\/h3>\n

GPT-4.1 (and its nano\/mini variants) lifts Azure\u2019s context ceiling to 1 million tokens<\/strong>, letting you pass entire codebases or multi-gigabyte corpora in one shot, while retaining GPT-4-class reasoning and function calling. That means fewer chunk-and-stitch hacks, simpler prompts, and major latency savings for large-document RAG or full-repo code reviews. Learn how to call it with the Responses API:<\/p>\n

Learn more<\/a><\/div><\/div>\n

GPT-image-1 Text-&-Image Generation<\/h3>\n

GPT-image-1 arrives in limited preview with sharper fidelity, reliable text rendering, editing \/ in-painting, and image-as-input support\u2014so you can build marketing creatives, design mocks, and visual KB answers directly in Foundry using the same REST patterns as DALLE 3.<\/p>\n

\"foundry<\/a>
Prompt: A bustling metallurgical foundry is portrayed in an isometric illustration. The scene features developers operating computers, typing into keyboards, speaking into microphones and overseeing an “ai agent factory” creation process. The environment is filled with datacenter elements such as long rows of server racks, neatly organized networking cables, and large cooling apparatuses.\u200b
Above the main workspace, a prominent banner stretches across the foundry floor, displaying the words “Happy Building” in bold, industrial-style lettering. The sign adds a touch of positivity and motivation amidst the intense industrial setting.<\/figcaption><\/figure><\/p>\n

Happy Building!<\/p>\n

Generate images<\/a><\/div><\/div>\n

o4-mini & o3 Reasoning Models<\/h3>\n

Need faster, cheaper reasoning? The new o-series pairs GPT-4-level logical depth with lower latency and aggressive pricing, making them ideal for agent planning, re-ranking, or embedded analytics where every millisecond (and penny) counts.<\/p>\n

\"A<\/a>
Source: https:\/\/openai.com\/index\/introducing-o3-and-o4-mini\/<\/figcaption><\/figure><\/p>\n
Learn more<\/a><\/div><\/div>\n

GPT-4o Audio (Transcribe & TTS)<\/h3>\n

gpt-4o-transcribe<\/code>, gpt-4o-mini-transcribe<\/code>, and gpt-4o-mini-tts<\/code> bring high-quality speech-to-text and controllable text-to-speech to Azure. Stream captions, build multilingual voice bots, or generate audio replies\u2014all via familiar \/audio<\/code> and \/realtime<\/code> endpoints.<\/p>\n

\"Demonstration<\/a>
Demonstration of the Azure OpenAI TTS Soundboard<\/figcaption><\/figure><\/p>\n

Get started<\/a><\/div><\/p>\n
\n

Agents<\/h2>\n

AI Red Teaming Agent (Preview)<\/h3>\n

Built atop Microsoft\u2019s PyRIT toolkit, this agent fires automated jailbreak and prompt-injection probes at your models, scores Attack Success Rate, and logs findings into Foundry dashboards\u2014making shift-left safety<\/em> a one-command reality.<\/p>\n

https:\/\/devblogs.microsoft.com\/foundry\/wp-content\/uploads\/sites\/89\/2025\/04\/AI-red-teaming-agent_final_blog_asset-1.mp4<\/a><\/video><\/div><\/p>\n

Learn more<\/a><\/div><\/p>\n

Semantic Kernel + A2A Interop<\/h3>\n

A new plug-in teaches Semantic Kernel to speak Google\u2019s Agent-to-Agent JSON-RPC protocol, enabling secure cross-cloud agent collaboration\u2014exchange context, not credentials, and orchestrate multi-modal workflows spanning Azure, GCP, and OSS runtimes.<\/p>\n

Learn more<\/a><\/div><\/p>\n

MCP Server Starter (Typescript)<\/h3>\n

Spin up an MCP-compliant<\/strong> server in minutes; the template wires Azure AI Agents to Claude Desktop (or any MCP client) via standard JSON messages\u2014no bespoke glue code required.<\/p>\n

Get started<\/a><\/div><\/p>\n

BYO Thread Storage + Monitor<\/h3>\n

Agent Service now lets you store conversation threads in your own Cosmos DB<\/strong> and surfaces run metrics in Azure Monitor\u2014boosting data residency compliance and giving SREs first-class observability out of the box.<\/p>\n

Quickstart<\/a><\/div><\/p>\n
\n

Tools<\/h2>\n

VS Code Foundry Extension<\/h3>\n

Test models, deploy agents, and copy sample code without leaving VS Code\u2014goodbye portal context-switching, hello faster inner loops.<\/p>\n

https:\/\/devblogs.microsoft.com\/foundry\/wp-content\/uploads\/sites\/89\/2025\/04\/25_66_Agent_v2.mp4<\/a><\/video><\/div><\/p>\n

Learn more<\/a><\/div><\/p>\n

Quality & Safety Evaluators<\/h3>\n

Four new quality metrics (intent-resolution, tool-call accuracy, task adherence, completeness) plus code-vulnerability and ungrounded-attribute safety checks plug straight into CI\/CD so every build ships with score gates.<\/p>\n

https:\/\/devblogs.microsoft.com\/foundry\/wp-content\/uploads\/sites\/89\/2026\/04\/eval-metrics-for-agents-2.mp4<\/a><\/video><\/div><\/p>\n

Learn more<\/a><\/div><\/p>\n

Fine-tuning support for GPT-4.1, GPT-4.1-mini, Phi-4 and Mistral models and more<\/h3>\n

The new\u00a0GPT-4.1 and GPT-4.1-mini<\/strong> models now support fine-tuning, offering enhanced reasoning and instruction-following capabilities, making them ideal for complex enterprise applications. Additionally, expanded serverless fine-tuning support for models like\u00a0Mistral<\/strong>,\u00a0Phi<\/strong>, and\u00a0NTT<\/strong> across all U.S. regions where base model inferencing is available,\u00a0improving latency and compliance with data residency requirements. The\u00a0Evaluation API<\/strong>\u00a0now supports code-first grading, allowing developers to score model outputs using built-in or custom graders, which simplifies A\/B testing, regression validation, and iterative model refinement.<\/p>\n

\"gpt<\/a><\/p>\n

Learn more<\/a><\/div><\/p>\n

Mem0 Persistent Memory Layer<\/h3>\n

Mem0 + Azure AI Search lets assistants remember<\/strong> user details across sessions via semantic retrieval\u2014boosting personalization without extra infra.<\/p>\n

Get started<\/a><\/div><\/p>\n

Content Understanding 2024-12-01 Preview<\/h3>\n

The new API adds generative & classification fields<\/strong>, faster video segmentation, and multi-analyzer docs, producing structured JSON ready for LLM ingestion across docs, audio, and video.<\/p>\n

Read the docs<\/a><\/div><\/p>\n

Document Intelligence v4.0 Container<\/h3>\n

Run the Layout model on-prem or at the edge via new v4.0 containers\u2014perfect for air-gapped PDF\/OCR scenarios that need local processing yet Azure-compatible APIs.<\/p>\n

Read the docs<\/a><\/div><\/p>\n
\n

Happy building\u2014and let us know what you ship with #AzureAIFoundry!<\/p>\n","protected":false},"excerpt":{"rendered":"

This post highlights the latest features and updates in Azure AI Foundry for April 2025, focusing on new models, agents, and tools that enhance AI development capabilities.<\/p>\n","protected":false},"author":185793,"featured_media":1563,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1,27],"tags":[3,29,12,4,2,28],"class_list":["post-312","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-microsoft-foundry","category-whats-new","tag-ai-development","tag-april-2025","tag-azure-openai","tag-generative-ai","tag-microsoft-foundry","tag-whats-new"],"acf":[],"blog_post_summary":"

This post highlights the latest features and updates in Azure AI Foundry for April 2025, focusing on new models, agents, and tools that enhance AI development capabilities.<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/posts\/312","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/users\/185793"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/comments?post=312"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/posts\/312\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/media\/1563"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/media?parent=312"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/categories?post=312"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/foundry\/wp-json\/wp\/v2\/tags?post=312"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}