Skip to main content
Back

Mistral Launches Agents API with Built-In Web Search, Code, MCP Tool

TRY NOW

May 28, 2025 3 min read

Hey,

Welcome to AlphaSignal – the most read newsletter by AI developers. 


We bring you the top 1% of news, papers, models, and repos, all summarized to keep you updated on the latest in AI.

IN TODAY'S SIGNAL

Read time: 3 min 54 sec

🎖️ Top News

› Mistral unveils Agents API that lets you run code, search the web, and generate images inside LLM workflows.

⇧ 5,385 Likes

📌 Assembly AI

› Customize speech recognition instantly with AssemblyAI's prompt-based model.

⚡️ Trending Signals

› Google DeepMind announces SignGemma to translate sign language into text.

⇧ 3,538 Likes

› Anthropic launches voice mode for Claude that will let you converse with the chatbot.

⇧ 3,439 Likes

› Google introduces Stitch to generate editable UI code from text or images.

⇧ 13,483 Likes

› Join SF’s top AI event: Greg Brockman, Cursor, and more. 30% off for readers.

From Our Sponsors

› Andrew Ng’s Landing AI upgrades agentic document extraction, processes PDFs in 8 seconds.

⇧ 3,749 Likes

  💻 Top Tutorials

› How to automate coding tasks using GitHub Copilot agent.

⇧ 5,329 Likes

› Build an AI agent to automate GitHub repos and PRs with Mistral.

☆ 1,712 Stars

› Watch Microsoft’s 1-hr course on AI agents with practical code.

⇧ 2,365,385 Likes

🧠 Python Tip

› Cache expensive ML function calls automatically


If you're enjoying AlphaSignal please forward this email to a colleague. 

It helps us keep this content free.

TOP NEWS

Agents

Mistral ships a new Agents API with support for tool use, orchestration, and external API calls

⇧ 5,385 Likes

The French AI company's new Agents API enables its LLMs to perform web search, execute code, and access external APIs. The most notable result: Mistral Large with web search achieves 75% accuracy on the SimpleQA benchmark, up from 23%.


Built-in Connectors

  • Web Search: Supports grounded citations from trusted sources in output

  • Code Execution: Runs Python in a secure, sandboxed environment

  • Image Generation: Generates images using FLUX1.1

  • Document Library: Supports retrieval-augmented generation using uploaded files

MCP Tools Integration

  • Attach APIs, internal systems, or structured data stores

  • Run external tasks directly from your agent workflows

  • Reuse tools across multiple agents with standardized access

Agent Orchestration

  • Assign tools and models to each agent for specific tasks

  • Define task handoffs to control which agents can delegate to others

  • Chain agent actions to handle complex tasks


Access and Availability

  • The Agents API is publicly available through Mistral’s developer portal

  • You can create agents by registering tools and assigning models

  • Access requires an API key issued via the developer dashboard

Community Feedback

Peri Carney

"Love how this makes Python development more fluid. Been looking for better code execution features like this."

Ronny Bruknapp
"This looks powerful for automating complex workflows. Are you planning to integrate this with your current tech stack? Curious about your use case."

Val Kamenski

"It’s interesting to see how model makers are bulking up with different APIs - evolving into AI-as-a-service platforms. It feels like it's not a race about who will have a better model now, but it's more about functionality for developers around these models."

Try now

Use AssemblyAI’s new voice model to get detailed audio analysis

Meet SLAM-1: The most powerful prompt-based Speech Language Model for real-world transcription that requires no manual training or complex configuration. 

Why SLAM-1 is the right choice for you:

  • Reduces missed key terms by 66% with prompt-based transcription.

  • Optimized for industries like healthcare, legal, finance, and tech

  • Improves user experience with high-quality transcripts

  • Understands industry-specific terms

  • Reduces post-processing costs and engineering effort

You can get started with the model within minutes through its easy-to-use API.

Try slam-1 today

partner with us

TOP TUTORIALS

Coding Assistance

How to automate coding tasks using GitHub Copilot agent

⇧ 5,329 Likes

Learn how Copilot’s agent mode handles multi-step coding tasks using natural language. Understand how it interacts with APIs, uses memory, and integrates with custom instructions to manage tasks beyond autocomplete.

GitHub Automation

Build an AI agent to automate GitHub repos and PRs with Mistral

☆ 1,712 Stars

This example demonstrates how to build a GitHub-integrated agent using Mistral Medium. The agent uses Mistral’s Agents API for persistent memory and tool orchestration, enabling multi-step workflows without manual scripting.

Agents

Microsoft’s 1-hr course on AI agents with practical code

⇧ 2,365,385 Likes

Learn core agent concepts in 10 short lessons that cover agentic RAG, design patterns, multi-agent systems, and production deployment. It includes hands-on code samples and a full beginner-friendly walkthrough

PYTHON TIP

Cache expensive ML function calls automatically

In ML workflows, repeated computations like feature calculations or data transformations can hinder the performance of your code. This function lets you automatically store and reuse results of expensive pure functions, making pipelines faster with zero extra logic.


When to use
Use it for deterministic, expensive functions like feature extraction, where inputs may repeat.

Benefits

Automatically handles caching with minimal code changes.


from functools import cache

@cache
def expensive_calc(x):
    return x**2 + 42

print(expensive_calc(10))  # Calculates
print(expensive_calc(10))  # Cached

At Alpha Signal, our mission is to build a sharp, engaged community focused on AI, machine learning, and cutting-edge language models, helping over 200,000 developers stay informed and ahead. We’re passionate about curating the best in AI, from top research and trending technical blogs to expert insights and tailored job opportunities. We keep you connected to the breakthroughs and discussions that matter, so you can stay in the loop without endless searching. We also work closely with partners who value the future of AI, including employers and advertisers who want to reach an audience as passionate about AI as we are.


Our partnerships are based on shared values of ethics, responsibility, and a commitment to building a better world through technology.Privacy is a priority at Alpha Signal. Our Privacy Policy clearly explains how we collect, store, and use your personal and non-personal information. By using our website, you accept these terms, which you can review on our website. This policy applies across all Alpha Signal pages, outlining your rights and how to contact us if you want to adjust the use of your information. We’re based in the United States. By using our site, you agree to be governed by U.S. laws.

Looking to promote your company, product, service, or event to 200,000+ AI developers? Let's work together.

Work with us