Mistral AI Unveils Magistral: Europe's First Multilingual AI Reasoning Models

Mistral releases a pair of AI reasoning models | TechCrunch

French AI lab Mistral is getting into the reasoning AI model game. On Tuesday morning, Mistral announced Magistral, its first family of reasoning models. Like other reasoning models -- e.g. OpenAI's o3 and Google's Gemini 2.5 Pro -- Magistral works through problems step-by-step for improved consistency and reliability across topics such as math and physics. Magistral comes in two flavors: Magistral Small and Magistral Medium. Magistral Small is 24 billion parameters in size, and is available for download from the AI dev platform Hugging Face under a permissive Apache 2.0 license. (Parameters are the internal components of a model that guide its behavior.) Magistral Medium, a more capable model, is in preview on Mistral's Le Chat chatbot platform and the company's API, as well as third-party partner clouds. "[Magistral is] suited for a wide range of enterprise use cases, from structured calculations and programmatic logic to decision trees and rule-based systems," writes Mistral in a blog post. "[The models are] fine-tuned for multi-step logic, improving interpretability and providing a traceable thought process in the user's language." Founded in 2023, Mistral is a frontier model lab building a range of AI-powered services, including the aforementioned Le Chat and mobile apps. It's backed by venture investors like General Catalyst, and has raised over €1.1 billion (roughly $1.24 billion) to date. Despite its formidable resources, Mistral has lagged behind other leading AI labs in certain areas, like developing reasoning models. Magistral doesn't appear to be an especially competitive release, either, judging by Mistral's own benchmarks. On GPQA Diamond and AIME, tests that evaluate a model's physics, math, and science skills, Magistral Medium underperforms Gemini 2.5 Pro and Anthropic's Claude Opus 4. Magistral Medium also fails to surpass Gemini 2.5 Pro on a popular programming benchmark, LiveCodeBench. Perhaps that's why Mistral touts Magistral's other strengths in its blog post. Magistral delivers answers at "10x" the speed of competitors in Le Chat, Mistral claims, and supports a wide array of languages, including Italian, Arabic, Russian, and Simplified Chinese. "Building on our flagship models, Magistral is designed for research, strategic planning, operational optimization, and data-driven decision making," the company writes in its post, "whether executing risk assessment and modelling with multiple factors, or calculating optimal delivery windows under constraints." The release of Magistral comes after Mistral debuted a "vibe coding" client, Mistral Code. A few weeks prior to that, Mistral launched several coding-focused models and rolled out Le Chat Enterprise, a corporate-focused chatbot service that offers tools like an AI agent builder and integrates Mistral's models with third-party services like Gmail and SharePoint.

France's Mistral launches Europe's first AI reasoning model

PARIS, June 10 (Reuters) - Mistral on Tuesday launched Europe's first AI reasoning model, which uses logical thinking to create a response, as it tries to keep pace with American and Chinese rivals at the forefront of AI development. The French startup has attempted to differentiate itself by championing its European roots, winning the support of French President Emmanuel Macron, as well as making some of its models open source in contrast to the proprietary offerings of OpenAI or Alphabet's (GOOGL.O), opens new tab Google. Mistral is considered Europe's best shot at having a home-grown AI competitor, but has lagged behind in terms of market share and revenue. Reasoning models use chain-of-thought techniques - a process that generates answers with intermediate reasoning abilities when solving complex problems. They could also be a promising path forward in advancing AI's capabilities as the traditional approach of building ever-bigger large language models by adding more data and computing power begins to hit limitations. For Mistral, which was valued by venture capitalists at $6.2 billion, an industry shift away from "scaling up" could give it a window to catch up against better capitalized rivals. China's DeepSeek broke through as a viable competitor in January through its low-cost, open-sourced AI models, including one for reasoning. OpenAI was the first to launch its reasoning models last year, followed by Google (GOOGL.O), opens new tab a few months later. Meta (META.O), opens new tab, which also offers its models open-sourced, has not yet released a standalone reasoning model, though it said its latest top-shelf model has reasoning capabilities. Mistral is launching an open-sourced Magistral Small model and a more powerful version called Magistral Medium for business customers. "The best human thinking isn't linear - it weaves through logic, insight, uncertainty, and discovery. Reasoning language models have enabled us to augment and delegate complex thinking and deep understanding to AI," Mistral said. American companies have mostly kept their most advanced models proprietary, though a handful, such as Meta, has released open-source models. In contrast, Chinese firms ranging from DeepSeek to Alibaba have taken the open-source path to demonstrate their technological capabilities. Mistral Small is available for download on Hugging Face's platform and can reason in languages including English, French, Spanish, Arabic and simplified Chinese. Reporting by GV De Clercq and Supantha Mukherjee; Editing by Stephen Coates Our Standards: The Thomson Reuters Trust Principles., opens new tab Suggested Topics:Artificial Intelligence Supantha Mukherjee Thomson Reuters Supantha leads the European Technology and Telecoms coverage, with a special focus on emerging technologies such as AI and 5G. He has been a journalist for about 18 years. He joined Reuters in 2006 and has covered a variety of beats ranging from financial sector to technology. He is based in Stockholm, Sweden. Kenrick Cai Thomson Reuters Kenrick Cai is a correspondent for Reuters based in San Francisco. He covers Google, its parent company Alphabet and artificial intelligence. Cai joined Reuters in 2024. He previously worked at Forbes magazine, where he was a staff writer covering venture capital and startups. He received a Best in Business award from the Society for Advancing Business Editing and Writing in 2023. He is a graduate of Duke University.

Mistral AI unveils Magistral reasoning model

AI startup Mistral AI has announced Magistral, a reasoning model geared to domain-specific, transparent, and multilingual reasoning. Announced June 10, Magistral is designed for research, strategic planning, operational optimization, and data-driven decision making, whether executing risk assessment and modeling with multiple factors, or calculating optimal delivery windows under constraints, Magistral AI said. Combining expertise across professional domains, transparent reasoning that can be followed and verified, and deep multilingual flexibility, Magistral is suited for enterprise use cases ranging from structured calculations and programmatic logic to decision trees and rule-based systems, according to the company. Magistral is fine-tuned for multi-step logic, improving interpretability, and providing a traceable thought process in the user's language, unlike general-purpose models, Mistral AI said. The company said the model excels in maintaining high-fidelity reasoning across numerous languages, and is particularly well-suited to reason in languages including English, French, Spanish, German, Italian, Arabic, Russian, and Simplified Chinese. It also enhances coding and development use cases, significantly improving project planning, back-end architecture, front-end design, and data engineering compared to non-reasoning models, according to the company.

Microsoft-backed AI lab Mistral is launching its first reasoning model in challenge to OpenAI

French founder of artificial intelligence startup Mistral AI, Arthur Mensch, attends the Viva Technology show at Parc des Expositions Porte de Versailles in Paris, France, on May 22, 2024. French artificial intelligence firm Mistral is on Tuesday launching its first reasoning model to compete with rival options from the likes of OpenAI and China's DeepSeek. The startup, which is backed by U.S. tech giant Microsoft, on Tuesday said that it plans to release its own reasoning model Magistral, which is "competitive with all the others," including OpenAI's o1 and Chinese AI firm DeepSeek's R1, according to CEO Arthur Mensch. Reasoning models are systems that can execute more complicated tasks through a step-by-step logical thought process. Mistral's new model "is great at mathematics [and] great at coding," Mensch told CNBC's Arjun Kharpal onstage during a fireside chat at London Tech Week. The Mistral boss said that the unique selling point of the company's upcoming Magistral reasoning model is that it'll be able to reason with European languages.

Exclusive: French AI startup Mistral unveils fast, multilingual reasoning models

Why it matters: Reasoning models aim to boost accuracy by using more computing power at inference time, a trend gaining steam amid diminishing returns from building ever-larger language models. Driving the news: Mistral is releasing two versions of its new Magistral family of reasoning models. Between the lines: Unlike many models that reason primarily in English, Mistral's can "think" in the query's native language, a potentially more efficient approach. The big picture: Based in France, Mistral has benefited from both its embrace of open source as well as the fact that it offers an alternative to the large American and Chinese tech companies. What they're saying: Anjney Midha from a16z, who led the company's Series A funding round and sits on its board, said that Mistral's focus on open source is its key selling point, even more than its unique geography.

Mistral's first reasoning model, Magistral, launches with large and small Apache 2.0 version

Join the event trusted by enterprise leaders for nearly two decades. VB Transform brings together the people building real enterprise AI strategy. Learn more European AI powerhouse Mistral today launched Magistral, a new family of large language models (LLMs) that marks the first from the company to enter the increasingly competitive space of "reasoning," or models that take time to reflect on their thinking to catch errors and solve more complex tasks than basic text-based LLMs. The announcement features a strategic dual release: a powerful, proprietary Magistral Medium for enterprise clients, and, notably, a 24-billion parameter open-source version, Magistral Small. The latter release appears calculated to reinforce the company's commitment to its foundational roots, following a period where it faced criticism for leaning into more closed, proprietary models such as its Medium 3 for enterprises, launched back in May 2025. A return to open source roots In a move that will undoubtedly be celebrated by developers and the wider AI community, Mistral is releasing Magistral Small under the permissive open source Apache 2.0 license. This is a crucial detail. Unlike more restrictive licenses, Apache 2.0 allows anyone to freely use, modify, and distribute the model's source code, even for commercial purposes. This empowers startups and established companies alike to build and deploy their own applications on top of Mistral's latest reasoning architecture without licensing fees or fear of vendor lock-in. This open approach is particularly significant given the context. While Mistral built its reputation on powerful open models, its recent release of Medium 3 as a purely proprietary offering drew concern from some quarters of the open-source community, who worried the company was drifting towards a more closed ecosystem, similar to competitors like OpenAI. The release of Magistral Small under such a permissive license serves as a powerful counter-narrative, reaffirming Mistral's dedication to arming the open community with cutting-edge tools. Competitive performance against formidable foes Mistral isn't just talking a big game; it came with receipts. The company released a suite of benchmarks pitting Magistral-Medium against its own predecessor, Mistral-Medium 3, and competitors from Deepseek. The results show a model that is fiercely competitive in the reasoning arena. On the AIME-24 mathematics benchmark, Magistral-Medium scores an impressive 73.6% on accuracy, neck-and-neck with its predecessor and significantly outperforming Deepseek's models. When using majority voting (a technique where the model generates multiple answers and the most common one is chosen), its performance on AIME-24 jumps to a staggering 90%. The new model also holds its own across other demanding tests, including GPQA Diamond, a graduate-level question-answering benchmark, and LiveCodeBench for coding challenges. While Deepseek-V3 shows strong performance on some benchmarks, Magistral-Medium consistently proves itself to be a top-tier reasoning model, validating Mistral's claims of its advanced capabilities. Enterprise power While Magistral Small caters to the open-source world, the benchmark-validated Magistral Medium is aimed squarely at the enterprise. It's acessible via Mistral's Le Chat interface and La Plateforme API, it delivers the top-tier performance needed for mission-critical tasks. Mistral is making this model available on major cloud platforms, including Amazon SageMaker, with Azure AI, IBM WatsonX, and Google Cloud Marketplace to follow. This dual-release strategy allows Mistral to have its cake and eat it too: fostering a vibrant ecosystem around its open models while monetizing its most powerful, performance-tested technology for corporate clients. Cost comparison When it comes to cost, Mistral is positioning Magistral Medium as a distinct, premium offering, even compared to its own models. At $2 per million input tokens and $5 per million output tokens, it represents a significant price increase from the older Mistral Medium 3, which costs just $0.40 for input and $2 for output. However, when placed against its external rivals, Magistral Medium's pricing strategy appears highly aggressive. Its input cost matches that of OpenAI's latest model and sits within the range of Gemini 2.5 Pro, yet its $5 output price substantially undercuts both, which are priced at $8 and upwards of $10, respectively. While it is considerably more expensive than specialized models like DeepSeek-Reasoner, it is an order of magnitude cheaper than Anthropic's flagship Claude Opus 4, making it a compelling value proposition for customers seeking state-of-the-art reasoning without paying the absolute highest market prices. Reasoning you can view, understand and use Mistral is pushing three core advantages with the Magistral line: transparency, multilingualism, and speed. Breaking away from the "black box" nature of many AI models, Magistral is designed to produce a traceable "chain-of-thought." This allows users to follow the model's logical path, a critical feature for high-stakes professional fields like law, finance, and healthcare, where conclusions must be verifiable. Furthermore, these reasoning capabilities are global. Mistral emphasizes the model's "multilingual dexterity," highlighting high-fidelity performance in languages including French, Spanish, German, Italian, Arabic, Russian, and Simplified Chinese. On the performance front, the company claims a major speed boost. A new "Think mode" and "Flash Answers" feature in Le Chat reportedly enables Magistral Medium to achieve up to 10 times the token throughput of competitors, facilitating real-time reasoning at a scale previously unseen. From code gen to creative strategy and beyond The applications for Magistral are vast. Mistral is targeting any use case that demands precision and structured thought, from financial modeling and legal analysis to software architecture and data engineering. The company even showcased the model's ability to generate a one-shot physics simulation, demonstrating its grasp of complex systems. But it's not all business. Mistral also recommends the model as a "creative companion" for writing and storytelling, capable of producing work that is either highly coherent or, as the company puts it, "delightfully eccentric." With Magistral, Mistral AI is making a strategic play to not just compete, but lead in the next frontier of AI. By re-engaging its open-source base with a powerful, permissively licensed model while simultaneously pushing the envelope on enterprise-grade performance, the company is signaling that the future of reasoning AI will be both powerful and, in a meaningful way, open to all.

Mistral AI debuts new Magistral series of reasoning LLMs - SiliconANGLE

The LLM series includes two algorithms on launch. The first, Magistral Small, is available under an open-source license and features 24 billion parameters. It's joined by a more capable, proprietary model called Magistral Small that will be available through Mistral AI's cloud services. Mistral AI is a Paris-based OpenAI competitor backed by more than $1 billion in funding. Alongside the newly launched reasoning-optimized models, it offers general-purpose LLMs and neural networks optimized for specialized tasks such as solving math problems. The launch of Magistral comes amid rumors that the company is seeking to raise another $1 billion from investors. The two models in the Magistral series share several features. Both understand multiple languages and ship with a chain-of-thought feature, which allows them to break down complex tasks into simpler sub-steps. Moreover, they can display the sub-steps involved in generating a prompt response, which enables users to verify its accuracy. Magistral Medium, Mistral's other new reasoning model, generates higher-quality output. The company compared it with Magistral Small by asking the models to solve problems from a qualifying exam for the 2024 U.S. Math Olympiad. Magistral Medium scored 73.6% with default settings and 90% with a configuration designed to boost output quality. Magistral Small scored 70.7% and 83.3%, respectively. Magistral Medium also includes speed optimizations not available in its open-source namesake. When users access the former model through Le Chat, Mistral's chatbot service, they can activate two settings called Think mode and Flash Answers. According to Mistral, the settings allow Magistral Medium to answer prompts nearly 10 times faster than competing models. In a paper accompanying the launch of Magistral, Mistral detailed how the LLM series was developed. The company used a popular AI training method known as reinforcement learning, or RL. "Instead of relying on existing implementations and RL traces distilled from prior models, we follow a ground up approach, relying solely on our own models and infrastructure," Mistral researchers wrote in the paper. The typical RL project involves two models: the LLM being trained and a so-called critic model that guides the training process by providing the LLM with feedback. According to Mistral, Magistral was trained using an RL method that removes the need for a critic model. This arrangement can improve the quality of LLM prompt responses. Mistral developed programs called generators and verifiers to manage the training process. Magistral used the generators to answer the practice questions in its training dataset. The verifiers, in turn, checked the accuracy of the model's answers. The trainers and verifiers spread the calculations involved in the workflow across a cluster of graphics cards. During the project, Mistral created several versions of the training workflow with which it trained Magistral and compared them. The company says that the test produced several new discoveries about RL. "We contribute insights that add to, or contradict, existing RLVR literature, for example on whether RL can improve upon the distillation SFT baseline for small models," the company's researchers wrote. Mistral's first discovery was that a version of Magistral trained solely on a coding dataset proved surprisingly adept at solving math problems. The opposite was true as well, the company determined. "The model demonstrates strong performance to out-of-domain tasks, showcasing the generalization ability of RL," Mistral's researchers wrote. The ability to apply knowledge from one field to another is important for many reasoning tasks. An earlier research paper observed that small models trained solely with RL can't compete with LLMs developed the same way. According to Mistral, its AI training tests showed that's not always the case. "We achieved strong results even with pure RL," the company's researchers detailed.

Mistral AI Unveils Its First Reasoning Model | AIM

French AI startup Mistral has unveiled Magistral, its first reasoning-focused language model. It aims to enhance transparent, multilingual, and domain-specific problem solving. Released in two variants, Magistral Small (24B, open-weight) and Magistral Medium (enterprise-grade), the model is designed for tasks that require step-by-step deliberation. Mistral says Magistral improves on the limitations of earlier models by offering more consistent reasoning in multiple languages and traceable logic across disciplines. This comes right after the release of its new enterprise-grade Document AI platform, which claims to set a new benchmark in speed and accuracy for OCR-based document processing. Magistral Medium scored 73.6% on the 2024 AIME benchmark and 90% with majority voting at 64-shot prompts. The open Magistral Small model scored 70.7% and 83.3%, respectively. The company mentioned that both models are tuned for legal research, financial modelling, software engineering, and regulated sectors like healthcare and government. "Magistral is fine-tuned for multi-step logic, improving interpretability and providing a traceable thought process in the user's language, unlike general-purpose models," the company wrote in the blog post. It supports reasoning in English, French, Arabic, German, Chinese, and several other languages. Mistral is also integrating the model into its Le Chat assistant, where a new 'Flash Answers' mode delivers responses at 10x the speed of competing systems. Magistral Medium is already accessible on La Plateforme and Amazon SageMaker, with support coming soon to IBM WatsonX, Azure AI, and Google Cloud. The company has released Magistral Small under Apache 2.0 on Hugging Face and shared a research paper detailing the model's training, infrastructure, and evaluation methodology. Mistral plans to iterate rapidly on the architecture, encouraging the developer community to build on its transparent reasoning framework.

Mistral's new reasoning model is a European first

Magistral is designed to 'think things through' in a manner similar to human beings, the company claims. French start-up Mistral AI has launched Europe's first AI reasoning model Magistral, which is designed for research, as well as domain-specific, transparent and multilingual reasoning and decision-making. AI reasoning models are large language models capable of performing complex tasks and breaking down a question, to offer a more nuanced and human-like response as it "thinks things through". Mistral was founded in France in 2023 by researchers Arthur Mensch, Guillaume Lample and Timothée Lacroix. The AI start-up aims to challenge the concept of 'Big AI' by democratising artificial intelligence and making it more accessible through open-source, efficient and innovative AI models, products and solutions. According to Mistral, "The best human thinking isn't linear - it weaves through logic, insight, uncertainty and discovery. Reasoning language models have enabled us to augment and delegate complex thinking and deep understanding to AI, improving our ability to work through problems requiring precise, step-by-step deliberation and analysis. "But this space is still nascent. Lack of specialised depth needed for domain-specific problems, limited transparency and inconsistent reasoning in the desired language, are just some of the known limitations of early thinking models." Magistral is a dual-release model that will be issued in two variants: Magistral Small, a 24B parameter open-source model; and Magistral Medium, a more powerful, enterprise version. The platform's chain-of-thought will work across a range of global languages and alphabets and it is suited to a number of use cases, such as structured calculations, programmatic logic, decision trees and rule-based systems. "As we've open-sourced Magistral Small, we welcome the community to examine, modify and build upon its architecture and reasoning processes to further accelerate the emergence of thinking language models." Last November, Mistral launched a new API for content moderation. This model is trained to classify text in a range of languages into one of nine categories: sexual, hate and discrimination, violence and threats, health, financial, law, dangerous and criminal content, self-harm, and personally identifiable information. In June last year, the start-up raised €600m in equity and debt financing at a valuation of €5.8bn. Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.

Mistral launches Magistral: A new reasoning-focused AI model family

Mistral, a French AI lab, announced on Tuesday the release of Magistral, a family of reasoning models designed to work through problems step-by-step, enhancing consistency and reliability across varied subjects. Magistral is available in two versions: Small and Medium. Magistral Small, a 24 billion parameter model, can be downloaded from Hugging Face under the Apache 2.0 license. Magistral Medium is accessible in preview on Mistral's Le Chat chatbot platform, the company's API, and partner clouds. According to Mistral's blog post, Magistral is suited for uses ranging from calculations and logic to decision trees and rule-based systems. The company adds that the models are fine-tuned for multi-step logic, thereby improving interpretability and providing a traceable thought process in the user's language. Mistral, founded in 2023, develops AI-powered services, including Le Chat and mobile apps. The company is backed by venture investors such as General Catalyst and has raised over €1.1 billion (approximately $1.24 billion) to date. According to Mistral's benchmarks, Magistral Medium underperforms compared to Google's Gemini 2.5 Pro and Anthropic's Claude Opus 4 on GPQA Diamond and AIME, tests measuring physics, math, and science skills. Magistral Medium also trails Gemini 2.5 Pro on LiveCodeBench, a programming benchmark. Mistral states in its blog post that Magistral delivers answers at "10x" the speed of competitors in Le Chat. The company also notes that the model supports multiple languages, including Italian, Arabic, Russian, and Simplified Chinese. Concerning the uses of Magistral, Mistral writes that it is designed for research, strategic planning, operational optimization, and data-driven decision-making. The company continues, stating it is capable of executing risk assessment and modeling along with calculating delivery windows under constraints. The Magistral release follows the debut of Mistral Code, a "vibe coding" client. Previously, Mistral introduced coding-focused models and Le Chat Enterprise, a corporate chatbot service that offers an AI agent builder and integrates Mistral's models with services such as Gmail and SharePoint.

Mistral's First-Ever Reasoning Models Are Finally Here

Magistral Small is a 24-billion-parameter AI model Both models are capable of multi-step reasoning Magistral Medium can be accessed via Amazon SageMaker Mistral released its first reasoning-focused artificial intelligence (AI) models on Tuesday. Dubbed Magistral, the large language model (LLM) is available in two variants -- Small and Medium. While Magistral Small is an open-source model, the Medium variant is an enterprise-focused closed model. Both models are capable of multi-step reasoning and show transparent chain-of-thought (CoT). The French AI startup highlighted that Magistral is designed for research, strategic planning, and data-driven decision making. A preview version of the reasoning model can also be tried out via the company's Le Chat platform. In a newsroom post, the Paris-based AI firm announced the release of the two Magistral models. Magistral Small, the open-weight LLM, is available to download and deploy via Mistral's Hugging Face listing with the Apache 2.0 licence. The model can be used for both academic and commercial use cases. On the other hand, Magistral Medium is a proprietary model which can be accessed via Amazon SageMaker. It will also be hosted on IBM WatsonX, Azure AI, and Google Cloud Marketplace soon. A preview version of the model can also be experienced in Le Chat or via API on La Plateforme. Coming to model details, the Magistral Small is a 24 billion parameter model. The parameters of the enterprise version have not been disclosed by the company. Focusing on Magistral Medium's capabilities, the model is claimed to have scored 73.6 percent on the AIME2024, comparable to DeepSeek-R1. On the other hand, the Small variant is said to have scored 70.7 percent, Mistral said. Both models can reason natively, and the CoT works in multiple global languages. The company said Magistral can reason in English, French, Spanish, German, Italian, Arabic, Russian, and Simplified Chinese languages. As with most reasoning models, Magistral can also handle structured calculations, programmatic logic, as well as decision trees and rule-based systems. Mistral says enterprises and users working in critical fields such as finance, healthcare, government, and law will get traceable reasoning within the models, allowing them to check the logical steps it took to reach a response. This will allow users to audit responses that are more sensitive than others, the company added.

France's Mistral unveils its first 'reasoning' AI model

French AI startup Mistral unveiled "Magistral," a new "reasoning" model designed to tackle complex problems with traceable logic, competing with models from OpenAI, Google, and Anthropic. This launch follows Apple's recent research paper, "The Illusion of Thinking," which questions the true reasoning capabilities of current AI models, suggesting fundamental limitations beyond certain complexity thresholds.French artificial intelligence startup Mistral on Tuesday announced a so-called "reasoning" model it said was capable of working through complex problems, following in the footsteps of top US developers. Available immediately on the company's platforms as well as the AI platform Hugging Face, the Magistral "is designed to think things through -- in ways familiar to us," Mistral said in a blog post. The AI was designed for "general purpose use requiring longer thought processing and better accuracy" than its previous generations of large language models (LLMs), the company added. Like other "reasoning" models, Magistral displays a so-called "chain of thought" that purports to show how the system is approaching a problem given to it in natural language. This means users in fields like law, finance, healthcare and government would receive "traceable reasoning that meets compliance requirements" as "every conclusion can be traced back through its logical steps", Mistral said. The company's claim gestures towards the challenge of so-called "interpretability" -- working out how AI systems arrive at a given response. Since they are "trained" on gigantic corpuses of data rather than directly programmed by humans, much behaviour by AI systems remains impenetrable even to their creators. Mistral also vaunted improved performance in software coding and creative writing by Magistral. Competing "reasoning" models include OpenAI's o3, some versions of Google's Gemini and Anthropic's Claude, or Chinese challenger DeepSeek's R1. The idea that AIs can "reason" was called into question this week by Apple -- the tech giant that has struggled to match achievements by leaders in the field. Several Apple researchers published a paper called "The Illusion of Thinking" that claimed to find "fundamental limitations in current models" which "fail to develop generalizable reasoning capabilities beyond certain complexity thresholds".

France's Mistral launches Europe's first AI reasoning model

PARIS (Reuters) -Mistral on Tuesday launched Europe's first AI reasoning model, which uses logical thinking to create a response, as it tries to keep pace with American and Chinese rivals at the forefront of AI development. The French startup has attempted to differentiate itself by championing its European roots, winning the support of French President Emmanuel Macron, as well as making some of its models open source in contrast to the proprietary offerings of OpenAI or Alphabet's Google. Mistral is considered Europe's best shot at having a home-grown AI competitor, but has lagged behind in terms of market share and revenue. Reasoning models use chain-of-thought techniques - a process that generates answers with intermediate reasoning abilities when solving complex problems. They could also be a promising path forward in advancing AI's capabilities as the traditional approach of building ever-bigger large language models by adding more data and computing power begins to hit limitations. For Mistral, which was valued by venture capitalists at $6.2 billion, an industry shift away from "scaling up" could give it a window to catch up against better capitalized rivals. China's DeepSeek broke through as a viable competitor in January through its low-cost, open-sourced AI models, including one for reasoning. OpenAI was the first to launch its reasoning models last year, followed by Google a few months later. Meta, which also offers its models open-sourced, has not yet released a standalone reasoning model, though it said its latest top-shelf model has reasoning capabilities. Mistral is launching an open-sourced Magistral Small model and a more powerful version called Magistral Medium for business customers. "The best human thinking isn't linear - it weaves through logic, insight, uncertainty, and discovery. Reasoning language models have enabled us to augment and delegate complex thinking and deep understanding to AI," Mistral said. American companies have mostly kept their most advanced models proprietary, though a handful, such as Meta, has released open-source models. In contrast, Chinese firms ranging from DeepSeek to Alibaba have taken the open-source path to demonstrate their technological capabilities. Mistral Small is available for download on Hugging Face's platform and can reason in languages including English, French, Spanish, Arabic and simplified Chinese. (Reporting by GV De Clercq and Supantha Mukherjee; Editing by Stephen Coates)