Mistral Large 3
Mistral Large 3 is Mistral AI’s flagship enterprise model, offered with strong European data sovereignty guarantees and competitive performance on multilingual tasks.
Released
February 12, 2026
Type
multimodal
License
proprietary
Context
256,000 tokens
Pricing
Input
$2.00 / 1M tokens
Output
$6.00 / 1M tokens
Capabilities
Links
Mistral Large 3 in the news
Financial Times · Jul 22, 2026
Samsung in talks to invest in Mistral at €20bn valuation
Samsung Electronics is in advanced discussions to invest as much as €1 billion in the French artificial intelligence startup Mistral AI. This potential investment would value the Paris-based company at approximately €20 billion, positioning Mistral AI as a significant European contender against dominant US tech firms. The talks, which have been ongoing for several weeks, highlight Samsung's strategic interest in bolstering its AI capabilities and securing access to cutting-edge AI technology. Mistral AI, founded in 2023 by former researchers from Meta and Google, has rapidly gained prominence for its open-source AI models. The company's valuation of €20 billion underscores the intense competition and significant capital flowing into the AI sector. A substantial investment from a global technology leader like Samsung could provide Mistral AI with the financial resources needed to accelerate its research and development, expand its talent pool, and compete more effectively on the international stage. For Samsung, this investment would represent a strategic move to enhance its AI offerings across its diverse product lines, including smartphones, home appliances, and semiconductors. The South Korean conglomerate has been actively pursuing AI integration, and a partnership with Mistral AI could grant it access to advanced AI models and expertise. The potential deal is part of a broader trend of major technology companies seeking to invest in or acquire promising AI startups to maintain a competitive edge in the rapidly evolving AI landscape.
MarkTechPost · Jul 14, 2026
Mistral AI Releases Robostral Navigate: An 8B Model Enabling Robots to Navigate Complex Environments Using a Single RGB Camera
Mistral AI released Robostral Navigate this week, marking its first model specifically designed for embodied navigation in robots. This 8 billion parameter model processes RGB images and plain-language instructions to guide a robot through complex environments, including offices, residential buildings, commercial spaces, and outdoor settings. A key feature of Robostral Navigate is its ability to achieve a 76.6% success rate on R2R-CE validation, navigating previously unseen environments using only a single RGB camera, without relying on depth sensors or LiDAR. The model operates by taking a single instruction, such as 'Leave the lobby, walk through the corridor, enter the supply room, and stop to face the second shelf,' and executing the entire task autonomously. This capability is particularly significant as it allows robots to navigate live spaces populated with people and unexpected obstacles, elements not present during its training phase. Traditional navigation systems often require multiple sensors or specialized hardware, whereas Robostral Navigate demonstrates efficiency and effectiveness with standard RGB cameras. Robostral Navigate employs a method called 'pointing' to determine its next movement. Given a task and its observation history, the model predicts the target's image coordinates within the current camera view and the desired orientation upon arrival. This pointing mechanism offers robustness against changes in camera intrinsics and world scale, unlike navigation commands based on metric displacements. When the target is outside the camera's field of view, the model defaults to using displacements within the robot's local coordinate frame, such as 'Move 2 meters forward, 1.5 meters to the left, and turn 25 degrees left.' This dual approach ensures continuous navigation even when visual cues are temporarily unavailable.
TechCrunch · Jul 4, 2026
What is Mistral AI? Everything to know about the OpenAI competitor
Mistral AI, established in 2023, has successfully raised significant funding rounds, positioning itself as a notable competitor in the artificial intelligence landscape. The company's stated mission is to "put frontier AI in the hands of everyone," indicating a commitment to making advanced AI technologies widely accessible. This ambition is supported by the substantial capital injections it has received since its inception. While specific funding amounts and investor details are not provided in the context of this summary, the repeated emphasis on "significant funding" suggests multiple successful fundraising efforts. These financial achievements are crucial for a company operating in the capital-intensive AI sector, which requires substantial investment in research, development, and talent acquisition. Mistral AI's strategy appears to involve offering open-source AI models, a move that can foster community adoption and rapid innovation, while also potentially generating revenue through enterprise solutions or specialized services. The company's rapid ascent since its founding in 2023 highlights the dynamic nature of the AI industry and the emergence of new players challenging established giants. By focusing on accessibility and open-source contributions, Mistral AI aims to carve out a distinct market position. The competitive landscape includes major players like OpenAI, which has also been a significant recipient of investment and is known for its advanced AI models. Mistral AI's approach suggests a strategy to leverage open-source principles to accelerate its growth and impact. The overarching goal of democratizing AI implies a long-term vision that extends beyond immediate commercial gains. It suggests a belief that broader access to powerful AI tools can unlock new opportunities across various industries and societal domains. The company's success in attracting significant investment underscores investor confidence in its leadership, technology, and market strategy. As Mistral AI continues to develop and deploy its AI models, its impact on the future of artificial intelligence is anticipated to grow.
MarkTechPost · Jul 3, 2026
Mistral AI Releases Leanstral 1.5: An Apache-2.0 Lean 4 Code Agent Model Solving 587 of 672 PutnamBench Problems
Mistral AI released Leanstral 1.5 this week, a code agent model specifically designed for the Lean 4 proof assistant. This new model targets applications in automated theorem proving and proof engineering, with its weights made available under the permissive Apache 2.0 license. A free API endpoint, leanstral-1-5, has also been launched. Leanstral 1.5 represents an update to the earlier Leanstral-2603 model and is part of Mistral AI's Mistral Small 4 family. The model employs a mixture-of-experts (MoE) architecture, activating a subset of specialized sub-networks for each token to maintain computational efficiency while leveraging a large overall capacity. It features 128 experts, with 4 active per token, a total size of 119 billion parameters, and 6.5 billion activated parameters per token. The model supports a context length of 256,000 tokens and can process both text and image inputs, though its output is limited to text. The training process for Leanstral 1.5 involved three stages: mid-training, supervised fine-tuning, and reinforcement learning using CISPO. Two distinct reinforcement learning environments were utilized to shape the model's agentic capabilities. In the multiturn environment, the model was tasked with proving or disproving theorems, iteratively refining its proofs based on feedback from the Lean compiler. The code agent environment immersed Leanstral within a simulated filesystem, enabling it to interact with files, execute bash commands, and utilize the Lean language server for real-time information on goals, errors, and types. This allowed for tasks such as completing partial proofs and creating auxiliary lemmas, with context compaction used to manage long tasks within the context window. Mistral AI reports that Leanstral 1.5 demonstrates strong performance, saturating the miniF2F benchmark and achieving 100% accuracy on its validation set. The model's correctness is verified against target theorems using Mistral's fork of SafeVerify. The model successfully solved 587 out of 672 problems on the PutnamBench, indicating significant advancements in its theorem-proving capabilities.
Fast Company · Jun 26, 2026
You know OpenAI and Nvidia. These are the AI companies building everything else
Mistral AI, a company founded in Paris in April 2023, is challenging the perception that frontier artificial intelligence development is solely concentrated in Silicon Valley. Established by former Google DeepMind employee Arthur Mensch, alongside Guillaume Lample and Timothée Lacroix from Meta, Mistral AI's initial premise was that advanced AI models could be developed without the massive compute resources typically associated with American hyperscalers. The company demonstrated this by releasing open-weight models, allowing widespread access for further development. Mistral AI reported substantial revenue growth, increasing from approximately $10 million in 2023 to over $400 million in annualized recurring revenue by early 2026. In September 2025, the company secured €1.7 billion in Series C funding, led by ASML, which valued Mistral AI at €11.7 billion. More recently, Mistral AI launched Vibe, a platform designed for research, drafting, and code deployment, functioning as an agentic system. This development highlights a broader trend of innovation occurring outside the traditional US-centric AI landscape, as exemplified by DeepSeek's market impact the previous year, which underscored that significant AI advancements are no longer limited to a few prominent companies and research institutions.