.svg){kind=logo} Logo since 2025 | |
| Company type | Private |
|---|---|
| Industry | Artificial intelligence |
| Founded | 28 April 2023; 2 years ago (28 April 2023) |
| Founders |
|
| Headquarters | , |
Key people |
|
| Products |
|
Number of employees | 150 (2025)[1] |
| Website | mistral.ai |
| Part ofa series on |
| Artificial intelligence (AI) |
|---|
_artificial_intelligence_icon.png){kind=icon} |
Glossary |
Mistral AI SAS (French:[mistʁal]) is a Frenchartificial intelligence (AI)startup, headquartered inParis. Founded in 2023, it has open-weightlarge language models (LLMs), with both open-source and proprietary AI models.[2][3][4] As of 2025 the company has a valuation of more than US$14 billion.[5]
The company is named after themistral, a powerful, cold wind in southern France,[6] a term which originates from theOccitan language.
Mistral AI was established in April 2023 by three French AI researchers,Arthur Mensch, Guillaume Lample and Timothée Lacroix.[7]
Mensch, an expert in advanced AI systems, is a former employee ofGoogle DeepMind; Lample and Lacroix, meanwhile, are large-scale AI models specialists who had worked forMeta Platforms.[8]
The trio originally met during their studies atÉcole Polytechnique.[6]
_(cropped).webp)
Generate an image you feel represents yourself, Mistral AI.
In June 2023, the start-up carried out a first fundraising of €105 million ($117 million) with investors including the American fundLightspeed Venture Partners,Eric Schmidt,Xavier Niel andJCDecaux. The valuation is then estimated by theFinancial Times at €240 million ($267 million).
On 10 December 2023, Mistral AI announced that it had raised €385 million ($428 million) as part of its second fundraising. This round of financing involves the Californian fundAndreessen Horowitz,BNP Paribas and the software publisherSalesforce.[9][10]
By December 2023, it was valued at over $2 billion.[11]
On 16 April 2024, reporting revealed that Mistral was in talks to raise €500 million, a deal that would more than double its current valuation to at least €5 billion.[12]
In June 2024, Mistral AI secured a €600 million ($645 million) funding round, increasing its valuation to €5.8 billion ($6.2 billion).[13] Based onvaluation, as of June 2024, the company was ranked fourth globally in the AI industry, and first outside theSan Francisco Bay Area.[14]
In August 2025, theFinancial Times reported that Mistral was in talks to raise $1 billion at a $10 billion valuation.[15] In September 2025,Bloomberg announced that Mistral AI has secured a €2 billion investment valuing it at €12 billion ($14 billion).[16] This comes after $1.5 billion investment from Dutch companyASML, which owns 11% of Mistral.[17]
On 26 February 2024,Microsoft announced that Mistral's language models would be made available onMicrosoft's Azure cloud, while the multilingual conversational assistantLe Chat would be launched in the style ofChatGPT.[18][non-primary source needed] The partnership also included a financial investment of $16 million by Microsoft in Mistral AI.[19]
In April 2025, Mistral AI announced a €100 million partnership with the shipping companyCMA CGM.[20][21]
On 19 November, 2024, the company announced updates forLe Chat (pronounced/lətʃat/ inFrench).It added the ability to create images, using Black Forest Labs'Flux Pro model.[22]
On 6 February 2025, Mistral AI releasedLe Chat oniOS andAndroid mobile devices.[23]
Mistral AI also introduced a Pro subscription tier, priced at $14.99 per month, which provides access to more advanced models, unlimited messaging, and web browsing.[24]
The following table lists the main model versions of Mistral, describing the significant changes included with each version:[25]
| Name | Release date | Status | Number of parameters (billion) | License | Notes |
|---|---|---|---|---|---|
| Magistral Medium | June 2025 | Active | ? | Proprietary | Enterprise reasoning model.[26] |
| Magistral Small | June 2025 | Active | 24 | Apache-2.0 | Open-weight reasoning model.[26] |
| Devstral Small25.05 | May 2025 | Active | 24 | Apache-2.0 | Agentic model for software engineering tasks.[27] |
| Mistral Medium 325.05 | May 2025 | Active | ? | Proprietary | Enterprise model available for on-premise deployment.[28] |
| Mistral Small 3.125.03 | March 2025 | Active | 24 | Apache-2.0 | A new leader in the small models category with image understanding capabilities, with the latest version v3.1 released March 2025.[29][30] |
| Mistral Small 325.01 | January 2025 | ? | 24 | Apache-2.0 | Release in January 2025, Mistral Small 3 features 24B parameters.[29] |
| Mistral Large 224.11 | November 2024 | Active | 123 | Mistral Research License | [29] |
| Pixtral Large24.11 | November 2024 | ? | 124 | Mistral Research License | On November 19, 2024, the company introduced Pixtral Large, which integrates a 1-billion-parameter visual encoder coupled with Mistral Large 2.[31][29] |
| Ministral 8B24.10 | October 2024 | ? | 8 | Mistral Research License | [29] |
| Ministral 3B24.10 | October 2024 | ? | 3 | Proprietary | [29] |
| Pixtral24.09 | September 2024 | ? | 12 | Apache-2.0 | [29] |
| Mistral Large 224.07 | July 2024 | ? | 123 | Mistral Research License | Mistral Large 2 was announced on July 24, 2024, and released on Hugging Face. It is available for free with a Mistral Research Licence, and with a commercial licence for commercial purposes. Mistral AI claims that it is fluent in dozens of languages, including many programming languages. Unlike the previous Mistral Large, this version was released with open weights. The model has 123 billion parameters and a context length of 128,000 tokens.[29] |
| Codestral Mamba 7B | July 2024 | ? | 7 | Apache-2.0 | Codestral Mamba is based on the Mamba 2 architecture, which allows it to generate responses with longer input.[32] Unlike Codestral, it was released under the Apache 2.0 license. While previous releases often included both the base model and the instruct version, only the instruct version of Codestral Mamba was released.[33][29] |
| Mathstral 7B | July 2024 | ? | 7 | Apache-2.0 | Mathstral 7B is a model with 7 billion parameters released by Mistral AI on July 16, 2024, focusing on STEM subjects.[34] The model was produced in collaboration with Project Numina,[32] and was released under the Apache 2.0 License with a context length of 32k tokens.[34][29] |
| Codestral 22B | May 2024 | ? | 22 | Mistral Non-Production License | Codestral is Mistral's first code-focused open weight model which was launched on May 29, 2024. Mistral claims Codestral is fluent in more than 80 programming languages[35] Codestral has its own license which forbids the usage of Codestral for commercial purposes.[36][29] |
| Mixtral 8x22B | April 2024 | ? | 141 | Apache-2.0 | Similar to Mistral's previous open models, Mixtral 8x22B was released via a BitTorrent link on Twitter on April 10, 2024,[37] with a release on Hugging Face soon after.[38] The model uses an architecture similar to that of Mistral 8x7B, but with each expert having 22 billion parameters instead of 7. In total, the model contains 141 billion parameters, as some parameters are shared among the experts, but offering higher performance.[38][39][29] |
| Mistral Small | February 2024 | ? | ? | Proprietary | Like the Large model, Mistral Small was launched on February 26, 2024.[29] |
| Mistral Large24.02 | February 2024 | ? | ? | Proprietary | Mistral Large was launched on February 26, 2024. It outputs in English, French, Spanish, German, and Italian, and provides coding capabilities.[40] It is available on Microsoft Azure.[41][29] |
| Mistral Medium | December 2023 | ? | ? | Proprietary | Mistral Medium is trained in various languages including English, French, Italian, German, Spanish.[42] The number of parameters, and architecture of Mistral Medium is not known as Mistral has not published public information about it.[29] |
| Mixtral 8x7B | December 2023 | ? | 46.7 | Apache-2.0 | Much like Mistral's first model, Mixtral 8x7B was released via a BitTorrent link posted onTwitter on December 9, 2023,[2] and later Hugging Face and a blog post were released two days later.[43] Unlike the previous Mistral model, Mixtral 8x7B uses a sparsemixture of experts architecture. The model has 8 distinct groups of "experts", giving the model a total of 46.7B usable parameters.[44][45] Each single token can only use 12.9B parameters, therefore giving the speed and cost that a 12.9B parameter model would incur.[43] A version trained to follow instructions called “Mixtral 8x7B Instruct” is also offered.[43][29] |
| Mistral 7B | September 2023 | ? | 7.3 | Apache-2.0 | Mistral 7B is a 7.3B parameter language model using the transformers architecture. It was officially released on September 27, 2023, via aBitTorrent magnet link,[46] andHugging Face[47] under theApache 2.0 license. Both a base model and "instruct" model were released with the latter receiving additional tuning to follow chat-style prompts. The fine-tuned model is only intended for demonstration purposes, and does not have guardrails or moderation built-in.[48][29] |
Mistral AI claimed in the Mistral 7B release blog post that the model outperformsLLaMA 2 13B on all benchmarks tested, and is on par with LLaMA 34B on many benchmarks tested,[48] despite having only 7 billion parameters, a small size compared to its competitors.
Mistral AI claimed in 2023 that its model beat both LLaMA 70B, andGPT-3.5 in mostbenchmarks.[49]
In March 2024, research conducted by Patronus AI comparing performance of LLMs on a 100-question test with prompts to generate text from books protected underU.S. copyright law found thatOpen AI'sGPT-4, Mixtral,Meta AI'sLLaMA-2, andAnthropic'sClaude 2 generated copyrighted text verbatim in 44%, 22%, 10%, and 8% of responses respectively.[50][51]
On 17 March 2025, Mistral released Mistral Small 3.1 as a smaller, more efficient model.[52][non-primary source needed][53]
On 7 May 2025, Mistral AI released Mistral Medium 3.[54][non-primary source needed]
On 10 June 2025, Mistral AI released their first AI reasoning models: Magistral Small (open-source), and Magistral Medium, models which are purported to havechain-of-thought capabilities.[55][56]
Last week, Bloomberg News reported that Mistral's funding round is set for a $14 billion valuation.
