Mistral Large

Mistral Large is the family of flagship large language models developed by Mistral AI, the Paris-based AI laboratory founded in 2023, and is the company's most capable general-purpose model line. The family comprises the original Mistral Large (released February 26, 2024, a closed API-only model scoring 81.2% on MMLU)^[1], Mistral Large 2 (July 24, 2024, a 123-billion-parameter dense model with a 128,000-token context window)^[2], Mistral Large 2.1 (November 18, 2024)^[3], and Mistral Large 3 (December 2, 2025, a 675-billion-parameter sparse mixture-of-experts model with 41 billion active parameters, released under Apache 2.0)^[4][5]. In October 2026 Mistral added Mistral Large 4, a natively multimodal model with about 1 trillion total parameters, first as a public preview on October 6 and then as open weights on October 29^[25][26]. Across these iterations the family has functioned as Mistral AI's premium offering, marketed as a European competitor to OpenAI's GPT-4 line, Anthropic's Claude 3.5 Sonnet, and Meta's Llama 3.1 405B for the most demanding workloads: multilingual document analysis, code generation, mathematical reasoning, and agentic orchestration.

The family's trajectory is notable for a steady move toward openness. Mistral Large 1 was a fully closed, API-only model. Mistral Large 2 and 2.1 added open weights under the restrictive Mistral Research License (MRL). Mistral Large 3, released on December 2, 2025, switched to a permissive Apache 2.0 license, the first time a flagship Mistral model has been released under terms that explicitly allow commercial deployment of the weights without a separate agreement^[4][5]. Together with the Microsoft distribution partnership announced alongside Mistral Large 1, which made Azure the first cloud platform to offer Mistral Large^[6], the family has been central to Mistral AI's reputation as Europe's most credible frontier-scale model developer.

Background

Mistral AI was founded in April 2023 by Arthur Mensch (formerly at Google DeepMind), Guillaume Lample and Timothée Lacroix (both formerly at Meta AI). Lample and Lacroix had been among the authors of the original LLaMA paper at Meta, and the new company was conceived as a European response to the perceived dominance of US firms in the foundation model market. Mistral's early product strategy rested on two pillars: efficient open-weight base models, and a closed flagship intended to generate commercial revenue.

The open-weight track moved quickly. Mistral 7B (September 2023) demonstrated that a 7-billion-parameter dense Transformer could outperform Llama 2 13B on every benchmark its paper evaluated, largely by combining grouped-query attention (GQA) with sliding window attention (SWA)^[7]. Mixtral 8x7B followed in December 2023, using a sparse mixture-of-experts (MoE) routing scheme that activated only ~12.9 billion of its 46.7 billion parameters per forward pass, delivering Llama 2 70B-class quality at a fraction of the inference cost^[8].

Those open releases established Mistral AI's technical credibility. The closed flagship track began two months later with the original Mistral Large, which was deliberately marketed not as a successor to Mixtral but as a premium API product positioned against GPT-4 and Claude 2, a different commercial proposition for a different customer.

The phrase "European OpenAI" became commonplace in coverage of the company through 2024. Mistral's investor base (including Andreessen Horowitz, Lightspeed, Nvidia, Salesforce, and Cisco) combined with its EU headquarters and French government endorsements, gave it strategic weight in European AI policy discussions, particularly around the EU AI Act^[9]. The Mistral Large family is the most direct expression of that "frontier-class but European" positioning.

What was Mistral Large 1 (February 2024)?

Mistral Large 1, identified internally as mistral-large-2402, was announced on February 26, 2024^[1]. The release came bundled with a distribution partnership with Microsoft that made Azure the first cloud platform to offer Mistral Large, through Azure AI Studio and Azure Machine Learning^[6]. Microsoft accompanied the announcement with a reported EUR 15 million (approximately $16.3 million at then-current exchange rates) investment in Mistral AI, structured to convert into equity at the company's next funding round^[10][11]. Mistral AI was valued at roughly EUR 2 billion at the time^[11].

At launch, Mistral Large 1 offered:

• A 32,000-token context window
• Native fluency in English, French, Spanish, German, and Italian
• Built-in function calling and a JSON output mode for structured generation
• Constrained output mode through La Plateforme for production reliability

Mistral AI described the model as "the world's second-ranked model generally available through an API (next to GPT-4)" on its launch blog^[1]. The headline benchmark was MMLU at 81.2%, which placed the model behind GPT-4 (then around 86% on MMLU) but ahead of Claude 2 and notably above Google's Gemini Pro. Mistral reported results on HellaSwag, WinoGrande, Arc Challenge, TriviaQA and TruthfulQA in English, and said the model "strongly outperforms LLaMA 2 70B on HellaSwag, Arc Challenge and MMLU benchmarks in French, German, Spanish and Italian"^[1].

Alongside Mistral Large, the same announcement introduced Mistral Small, a lower-latency model positioned below Mistral Large in capability but above Mixtral 8x7B in performance, and the first version of Le Chat, Mistral's consumer-facing assistant. Together with Mixtral 8x7B and Mistral 7B, these formed Mistral's first explicit tier structure: open weights at the small/medium scale, closed commercial flagships at the top.

Mistral Large 1 was a fully proprietary closed model. Its weights were never released publicly, its architecture and parameter count were not disclosed, and access was restricted to La Plateforme and Azure. That deliberate closure reflected Mistral's stated need to establish a commercial revenue stream alongside its open research models^[1].

Initial launch pricing on La Plateforme was set at $8 per million input tokens and $24 per million output tokens, about 20% below contemporary GPT-4 Turbo pricing of $10 and $30^[28]. Mistral later adjusted this downward, with the 2402 variant priced at $4 input / $12 output by mid-2024 as the Mistral Large 2 release approached.

The Microsoft partnership drew immediate regulatory attention. Within 48 hours of the announcement, the European Commission stated it would examine the investment as part of its broader scrutiny of partnerships between large cloud platforms and generative AI startups, prompted in part by the OpenAI-Microsoft precedent^[12][13]. The Commission concluded its initial review without action, but the episode foreshadowed continuing tension between European AI policy and US cloud-platform partnerships through 2024 and 2025.

What is Mistral Large 2 (July 2024)?

Mistral Large 2, identified as mistral-large-2407, was announced on July 24, 2024 under the blog post title Large Enough^[2]. It was a substantially larger and more capable model than its predecessor across every measurable dimension, and crucially, it shipped with downloadable weights for the first time in the Large family.

The headline specifications were:

• Total parameters: 123 billion (dense, all active per forward pass)
• Context window: 128,000 tokens (up from 32,000 in Large 1)
• License: Mistral Research License (MRL)
• Architecture: Standard decoder-only Transformer with grouped-query attention^[14]

The design goal, as stated by Mistral AI, was a model "designed for single-node inference with long-context applications in mind: its size of 123 billion parameters allows it to run at large throughput on a single node"^[2]. This pragmatic constraint placed an upper bound on parameter count: 123B fits in bf16 across eight 80GB GPUs with room for KV cache, while 405B (the size of Llama 3.1 405B, released one day earlier on July 23, 2024) requires significantly more hardware.

At launch, Guillaume Lample (Mistral's chief scientist) summarized the comparison on X (formerly Twitter): "On many benchmarks (notably in code generation and math), it is superior or on par with Llama 3.1 405B"^[15]. Independent benchmarking through August 2024 broadly supported this claim on coding benchmarks while showing Llama 3.1 405B retaining an edge on the most knowledge-heavy evaluations.

Architecture details

While Mistral AI did not publish a formal arXiv technical report for Mistral Large 2, a notable departure from the Mistral 7B paper, key architectural parameters can be derived from the Hugging Face configuration files and third-party analysis^[14]:

Mistral Large 2 uses grouped-query attention with a ratio of 12:1 (96 query heads sharing across 8 KV heads), which significantly reduces KV cache memory pressure during long-context inference. The published configuration sets the RoPE base frequency (rope_theta) to 1,000,000 with 131,072 maximum positions^[23].

Multilingual capabilities

Mistral Large 2 was trained with explicit emphasis on multilingual coverage. The model card and launch post together list English, French, German, Spanish, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Arabic, Hindi, and Polish^[2][14]. For code, it covers 80+ programming languages with particular strength noted in Python, Java, C, C++, JavaScript, Bash, Swift, and Fortran.

Multilingual MMLU scores at launch^[2][14]:

The Korean score (60.1%) is the lowest in the table.

Behavioral improvements

Mistral AI made two explicit behavioral claims that distinguished Mistral Large 2 from its predecessor^[2]:

• Reduced hallucination through uncertainty acknowledgment. The model was trained to respond with explicit "I don't know" statements when its internal confidence was low, rather than generating plausible-sounding but incorrect answers.

• Improved function calling. The model gained support for both parallel tool calls (multiple tools invoked simultaneously) and sequential tool chains (one tool's output feeding the next), according to Mistral^[2].

Benchmarks at launch

Benchmark scores reported on the Mistral AI model card and launch blog^[2][14]:

Deployment and pricing

Weights were released on Hugging Face under mistralai/Mistral-Large-Instruct-2407^[14]. Inference at bf16 precision requires over 300 GB of GPU memory (typically eight 80GB GPUs such as H100 or A100), or approximately 75 GB at fp4 quantization. The OVHcloud blog published one of the first public reference architectures for sovereign self-hosted deployment, demonstrating it on a single 8x H100 node^[16].

API pricing at launch was set at $3.00 per million input tokens and $9.00 per million output tokens. This represented a roughly 25% reduction relative to Mistral Large 1's late-life pricing, and was reduced further to $2.00 input / $6.00 output on September 17, 2024^[29].

What changed in Mistral Large 2.1 (November 2024)?

Mistral Large 2.1, identified as mistral-large-2411, was released on November 18, 2024^[3]. It is best understood as a maintenance refresh rather than a new model generation: the 123-billion-parameter architecture and 128,000-token context window are unchanged from Mistral Large 2. Training focused on three areas that had drawn the most enterprise feedback after the July 2024 release: long context reliability, function calling, and system prompt adherence^[3].

The most user-visible change was a new instruction template, v7, which introduced explicit system prompt delimiters:

<s>[SYSTEM_PROMPT] <system prompt>[/SYSTEM_PROMPT][INST] <user message>[/INST] <assistant response></s>[INST] <user message>[/INST]

Previous Mistral templates lacked a dedicated system prompt slot, which made it difficult to reliably distinguish system-level instructions from user input in multi-turn agentic deployments. The v7 template addressed this directly, and Mistral AI recommended that even minimal system prompts be included for best results^[3].

Function calling reliability improvements were the second major change. Mistral described function calling as improved over 2407 but published no detailed figures^[23].

The 2411 release coincided with the announcement of Pixtral Large, a multimodal variant that pairs Mistral Large 2's text backbone with a vision encoder. Pixtral Large is documented separately; Mistral describes it as a 123B multimodal decoder with a 1B-parameter vision encoder that extends Mistral Large 2^[30].

Pricing remained at $2.00 per million input / $6.00 per million output. The 2411 variant was added across the major distribution platforms: La Plateforme, Azure AI Foundry, Amazon Bedrock, Google Cloud Vertex AI (where it was called Mistral Large 24.11)^[17], and IBM watsonx, through November and December 2024.

What is Mistral Large 3 (December 2025)?

Mistral Large 3, identified as mistral-large-2512 (or mistral-large-latest on the API), was announced on December 2, 2025 as part of the broader Mistral 3 release, which also introduced the Ministral 3 family of small models (3B, 8B, and 14B dense variants)^[4][18]. The release marked the most significant architectural shift in the Mistral Large family to date, abandoning the dense Transformer used in Large 1, 2, and 2.1 in favor of a sparse mixture-of-experts (MoE) design, and switching the license to Apache 2.0^[5].

Architecture

Mistral Large 3 uses a granular sparse mixture-of-experts architecture with the following published characteristics^[5]:

• Total parameters: 675 billion
• Active parameters per token: 41 billion
• Decomposition: 673B granular MoE language model + 2.5B vision encoder
• Context window: 256,000 tokens (up from 128,000 in Large 2/2.1)
• License: Apache 2.0
• Modality: Text + image input, text output
• Training infrastructure: 3,000 NVIDIA H200 GPUs^[4]

The "granular" qualifier refers to a finer-grained expert decomposition than the original Mixtral models; Mixtral 8x7B used eight experts with two active per token, while Mistral Large 3 uses a substantially larger expert pool with more experts active per token, producing smoother routing behavior and reduced load-balancing overhead. The configuration published with the weights lists 128 routed experts, 4 of them active per token, plus one shared expert^[5].

The compute-vs-memory tradeoff of MoE is the central technical motivation for the architectural switch. With 41B active parameters, inference compute per token is comparable to a 41B dense model, while the 675B total parameter budget provides representational capacity closer to dense models several times larger. This significantly improves the throughput-quality frontier compared to the 123B dense Mistral Large 2, which had to activate every parameter for every token.

The 2.5B vision encoder makes Mistral Large 3 natively multimodal at the flagship tier for the first time in the Mistral Large family. Previous multimodal capability had required Pixtral Large as a separate model.

Is Mistral Large 3 open source?

Apache 2.0 is a significant departure from the Mistral Research License used for Mistral Large 2 and 2.1. Under Apache 2.0^[5]:

• Commercial deployment of the weights is permitted without a separate license
• Modifications and derivative works are permitted
• The weights can be redistributed under the same license
• Patents implicated by use of the weights are granted to users

This places Mistral Large 3 alongside Llama 3.1 (under Meta's more restrictive custom Llama community license), Qwen3, and DeepSeek V3, making the entire frontier-scale open-weight tier governed by relatively permissive terms by late 2025.

Benchmark performance

Public benchmark scores for Mistral Large 3^[20][21]:

On the LMArena leaderboard at launch, Mistral Large 3 debuted at #2 among open-source non-reasoning models and #6 overall among open-source models^[4]. Artificial Analysis placed it slightly above the median (22) for open-weight non-reasoning models of similar size on intelligence in March 2026, with a very competitive time to first token of about 1.0 second but a below-average output speed of about 48 tokens per second^[21].

Mistral AI has been candid about the model's positioning: Mistral Large 3 is not a dedicated reasoning model and is explicitly outperformed by reasoning-specialized models (such as those from the o-series, DeepSeek R1 lineage, or Mistral's own ministral-14B reasoning variant which scores 85% on AIME '25) on extended reasoning tasks^[4][5]. The model is targeted instead at the general-purpose enterprise workload: multilingual content generation, code generation, document analysis, RAG pipelines, and agentic orchestration, where breadth of capability matters more than depth on hard reasoning benchmarks.

Deployment

Mistral Large 3 was made available at launch through Mistral AI Studio (La Plateforme), Hugging Face (under mistralai/Mistral-Large-3-675B-Instruct-2512), Microsoft Azure AI Foundry, Amazon Bedrock, IBM watsonx, and inference providers including OpenRouter, Fireworks and Together AI, with NVIDIA NIM and AWS SageMaker announced as coming soon^[4][18]. NVIDIA's announcement emphasized optimizations for GB200 NVL72 systems, claiming a 10x performance improvement compared to the prior generation H200 hardware on which the model was trained^[22].

Deployment precision options on Hugging Face^[5]:

• FP8: Single node of B200s or H200s
• NVFP4: Single node of H100s or A100s (separate quantized model card)
• BF16: Available as a separate model card; requires multi-node deployment

API pricing on La Plateforme at launch: $0.50 per million input tokens / $1.50 per million output tokens, a substantial reduction from the $2.00/$6.00 of Mistral Large 2.1, reflecting the lower active-parameter compute cost of the MoE architecture.

What is Mistral Large 4 (October 2026)?

Mistral opened a public preview of Mistral Large 4 on October 6, 2026, nicknaming it "le Chonk." The company described it as a natively multimodal model with 1 trillion parameters, 52 billion of them active, and as its largest and most capable model to date. Mistral said it was trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in its own data centers in Europe, and offered the preview through an API on Mistral Studio^[25].

On October 29, 2026, Mistral released the weights of an improved version on Hugging Face as mistralai/Mistral-Large-4-1T-A52B. The license is the Mistral Ecosystem Open-Weight license, which Mistral describes as a modified MIT license: the model is free to use, modify, fine-tune, and commercialize, but businesses that operate a Model-as-a-Service offering must obtain a separate commercial license before using it commercially. Mistral priced the API at $0.70 per million input tokens and $2.10 per million output tokens^[26].

How does the architecture compare across the family?

The Mistral Large family spans two distinct architectural eras:

The Large 2 to Large 3 transition is the most significant architectural break: from a dense Transformer with all parameters active per token, to a granular MoE with sparse expert routing; from multimodality through a separate model (via Pixtral) to integrated multimodality; and from research-only weight access to fully permissive commercial open weights.

How is Mistral Large licensed?

Through Mistral Large 3, the Mistral Large family used three distinct licensing regimes (Mistral Large 4 uses a fourth, described above):

Mistral Large 1 (proprietary, closed). No weight release. Access only through La Plateforme and Azure. Public self-hosting was not possible; Mistral's launch post offered self-deployment with weight access only by arrangement^[1]. This was an explicit commercial-revenue strategy at the time of the original Microsoft partnership^[1].

Mistral Large 2 and 2.1 (Mistral Research License). The MRL allows downloading, modifying, and using the weights for research and non-commercial purposes. Commercial deployment (any use that involves charging for access, embedding the model in a commercial product, or serving the model to third-party users) requires a separate commercial license obtained directly from Mistral AI^[14]. The MRL's restrictions also extend to the model's outputs, which may likewise be used only for research purposes^[14]. API access through La Plateforme, Azure, AWS, Vertex AI, and IBM watsonx constitutes commercial licensing through Mistral's distribution agreements, so API users do not need to negotiate separately.

Mistral Large 3 (Apache 2.0). Permissive open-source license permitting commercial use, modification, redistribution, and patent grant^[5]. This brings Mistral Large 3 in line with the broader frontier open-weight ecosystem (Llama 3, Qwen3, DeepSeek V3).

The licensing trajectory across the family (proprietary, then research-only open weights, then Apache 2.0) mirrors a broader industry shift in late 2024 and 2025 toward more permissive terms for flagship-tier open weights, driven by competitive pressure from Chinese labs (DeepSeek, Qwen, Yi) releasing comparably-sized models under permissive licenses.

Where can you use Mistral Large?

Mistral Large models have historically been distributed through multiple channels:

La Plateforme (console.mistral.ai) is Mistral AI's direct API service. It hosts Mistral's current models with pay-as-you-go pricing and fine-tuning capabilities. Mistral Large 3 was the flagship there from December 2025; on October 6, 2026, Mistral opened a preview API for Mistral Large 4 on Mistral Studio^[25].

Microsoft Azure was the first distribution partner for Mistral Large 1 in February 2024, and Azure AI Foundry continues to host Mistral Large 3 (along with the deprecated 2411 and earlier 2407 versions during their respective deprecation windows)^[6].

Amazon Bedrock and SageMaker added Mistral Large 2 in 2024 and host Mistral Large 3 as of December 2025. AWS published a reference architecture for agentic RAG pipelines using Mistral Large 2 with LlamaIndex during the 2024-2025 period.

Google Cloud Vertex AI added Mistral Large 24.11 alongside Codestral 25.01^[17].

IBM watsonx offered Mistral Large 2 starting in 2024

Le Chat was Mistral's consumer-facing assistant. Mistral has since renamed Le Chat as Vibe, which its help center describes as a unified agent for professional productivity and coding with Work, Code, and Chat modes^[27]. When it released the Large 4 weights, Mistral invited users to try the model in Vibe Code^[26].

Hugging Face hosts the open weights: mistralai/Mistral-Large-Instruct-2407 (MRL), mistralai/Mistral-Large-Instruct-2411 (MRL), and mistralai/Mistral-Large-3-675B-Instruct-2512 (Apache 2.0)^[14][23][5].

Deprecation status (November 2026)

Mistral Large 2.1 was deprecated on February 27, 2026 and retired on May 31, 2026. Mistral Large 3 was deprecated on October 31, 2026, with retirement scheduled for April 30, 2027 and Mistral Large 4 named as its replacement^[3].

How does Mistral Large 3 compare to competitors?

The competitive landscape Mistral Large 3 entered in December 2025 is denser than the one Mistral Large 1 entered in early 2024. The table below compares it with the 2024-era flagships the Mistral Large family was originally positioned against (competitor figures are from each vendor's own launch materials):

Mistral Large 3's distinguishing features at launch were the combination of permissive Apache 2.0 licensing, integrated image input, and a 256k context window.

For enterprises evaluating Mistral Large 3 against closed alternatives, the practical differentiator is self-hostability. GPT-4o and Claude 3.5 Sonnet are not available for on-premises deployment at any price, while Mistral Large 3 can be downloaded and run inside the enterprise security perimeter, a hard requirement for regulated industries such as finance, healthcare, and defense, and a soft preference for many EU-based organizations facing GDPR and data sovereignty constraints.

What is Mistral Large used for?

Across the Mistral Large family, four use case categories have dominated enterprise deployment:

Multilingual document processing. Mistral Large's first-party support for French, German, Spanish, Italian, Portuguese, and Dutch at scores above 80% on multilingual MMLU has made it a default choice for European enterprises with multi-jurisdictional document flows. Insurance claims processing, contract review, and KYC pipelines are commonly cited deployments.

Retrieval-augmented generation. The combination of a long context window (128k in Large 2/2.1, 256k in Large 3) with the model's explicit uncertainty-acknowledgment behavior makes it well-suited to RAG pipelines where precise citation from retrieved sources matters more than free-form generation. Mistral AI promoted this behavior explicitly in the Large 2 launch^[2].

Code generation. Mistral Large 2's HumanEval score of 92% and coverage of 80+ programming languages established it as a credible alternative to GPT-4o and Claude 3.5 Sonnet for code tasks. In practice, many deployments pair Mistral Large with Codestral (Mistral's specialized 22B code model), using Large for architecture-level reasoning and Codestral for high-throughput inline completions.

Agentic orchestration. Function calling reliability, strengthened in Large 2.1's v7 template and further in Large 3, has positioned the family as the orchestration model in multi-step agent pipelines.

Reception and legacy

Each release in the Mistral Large family was received in the context of an evolving competitive landscape:

Mistral Large 1 (February 2024) drew most of its attention from the Microsoft partnership. Coverage focused less on the model's technical capability and more on what it signaled: that Microsoft was hedging its OpenAI bet, that Europe had a frontier-scale model lab worth funding, and that the regulatory politics of AI partnerships were now actively contested in Brussels^[11][12].

Mistral Large 2 (July 2024) was the release that earned Mistral AI broad technical credibility. Releasing 123B-parameter open weights one day after Llama 3.1 405B's release, and posting competitive code generation and math benchmarks despite being one-third the size, established the model as a serious contender in the open-weight flagship tier^[15]. IBM's decision to offer Mistral Large 2 under its own customer license through watsonx was widely read as a signal that regulated-industry enterprises now considered Mistral a tier-one vendor.

Mistral Large 2.1 (November 2024) received more measured coverage, in part because it was a refinement rather than a new generation. The v7 instruction template was generally well-received, and the function calling improvements addressed real production pain points. The simultaneous release of Pixtral Large reinforced Mistral's commitment to the 123B family at the close of 2024.

Mistral Large 3 (December 2025) was Mistral AI's most ambitious release to date and the company's first frontier-scale MoE flagship. The Apache 2.0 license drew the most attention in early coverage, as it placed Mistral on equal footing with Meta's permissive Llama community license and DeepSeek's MIT release. Reception of the model itself was more nuanced: while the throughput-quality tradeoff of MoE was widely praised, the model's middle-of-the-pack position on Artificial Analysis's Intelligence Index left observers noting that Mistral had matched DeepSeek V3's architectural class without clearly surpassing it on benchmarks^[21].

Criticism of the family across its history has clustered around several recurring themes. The Mistral Research License for Large 2 and 2.1 drew complaints from developers accustomed to the more permissive Llama community license, particularly around its research-only restriction, which also covers model outputs. The model's inference speed at scale was slower than competitors with similar parameter counts during the 123B dense era, partly because the 8-GPU single-node requirement was less amenable to high-throughput batching than smaller models. The absence of a formal arXiv technical report for Mistral Large 2 made independent architectural analysis more difficult than was the case for Mistral 7B^[7].

The legacy of the Mistral Large family within Mistral AI's broader product line is the establishment of a credible commercial flagship tier alongside the open research models, and the demonstration that a European lab could produce frontier-scale models without US-only dependencies. The transition from MRL to Apache 2.0 with Large 3 closes one chapter of that strategy: the closed commercial flagship has been replaced with an open-weight commercial flagship, and revenue capture has shifted from license fees to managed-API pricing and enterprise contracts.

References

1. ^1 ^2 ^3 ^4 ^5 ^6 ^7Mistral AI. "Au Large." February 26, 2024. mistral.ai/...mistral-large
2. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9Mistral AI. "Large Enough." July 24, 2024. mistral.ai/...mistral-large-2407
3. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8Mistral AI Documentation. "Mistral Large 2.1 (24.11)." docs.mistral.ai/...models
4. ^1 ^2 ^3 ^4 ^5 ^6 ^7Mistral AI. "Introducing Mistral 3." December 2, 2025. mistral.ai/...mistral-3
5. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10Hugging Face. "mistralai/Mistral-Large-3-675B-Instruct-2512." huggingface.co/...Mistral-Large-3-675B-Instruct-2512
6. ^1 ^2 ^3Microsoft Azure Blog. "Microsoft and Mistral AI announce new partnership to accelerate AI innovation and introduce Mistral Large first on Azure." February 26, 2024. azure.microsoft.com/...istral-large-first-on-azure
7. ^1 ^2Jiang, Albert Q. et al. "Mistral 7B." arXiv:2310.06825. October 2023. arxiv.org/...2310.06825
8. ^Jiang, Albert Q. et al. "Mixtral of Experts." arXiv:2401.04088. January 2024. arxiv.org/...2401.04088
9. ^Reuters. "Microsoft's deal with Mistral AI faces EU scrutiny." February 27, 2024. reuters.com/...ral-ai-faces-eu-scrutiny-2024-02-27
10. ^TechCrunch. "Microsoft made a $16M investment in Mistral AI." February 27, 2024. techcrunch.com/...million-investment-in-mistral-ai
11. ^1 ^2 ^3CNBC. "Microsoft invests in Europe's Mistral AI to expand beyond OpenAI." February 26, 2024. cnbc.com/...pes-mistral-ai-to-expand-beyond-openai
12. ^1 ^2Bloomberg. "Microsoft's Mistral AI Investment to Be Examined by EU Watchdog." February 27, 2024. bloomberg.com/...ent-to-be-examined-by-eu-watchdog
13. ^The Register. "EU probes Microsoft's EUR 15M stake in AI startup Mistral." February 28, 2024. theregister.com/...eu_microsoft_mistral
14. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9Hugging Face. "mistralai/Mistral-Large-Instruct-2407." huggingface.co/...Mistral-Large-Instruct-2407
15. ^1 ^2Lample, Guillaume (@GuillaumeLample). X post announcing Mistral Large 2. July 24, 2024. x.com/...1816135838448972240
16. ^OVHcloud Blog. "Reference Architecture: deploying the Mistral Large 123B model in a sovereign environment." 2024. blog.ovhcloud.com
17. ^1 ^2Google Cloud Blog. "Announcing Mistral AI's Mistral Large 24.11 and Codestral 25.01 models on Vertex AI." 2025. cloud.google.com/...rge-and-codestral-on-vertex-ai
18. ^1 ^2NVIDIA Blog. "NVIDIA Partners With Mistral AI to Accelerate New Family of Open Models." December 2, 2025. blogs.nvidia.com/...mistral-frontier-open-models
19. IntuitionLabs. "Mistral Large 3: An Open-Source MoE LLM Explained." 2025. intuitionlabs.ai/...tral-large-3-moe-llm-explained
20. ^LLM Stats. "Mistral Large 3 (675B Instruct 2512) Benchmarks, Pricing & Context Window." llm-stats.com/...mistral-large-latest
21. ^1 ^2 ^3Artificial Analysis. "Mistral Large 3 - Intelligence, Performance & Price Analysis." artificialanalysis.ai/...mistral-large-3
22. ^NVIDIA Blog. "NVIDIA Partners With Mistral AI to Accelerate New Family of Open Models." December 2, 2025. blogs.nvidia.com/...mistral-frontier-open-models
23. ^1 ^2 ^3Hugging Face. "mistralai/Mistral-Large-Instruct-2411." huggingface.co/...Mistral-Large-Instruct-2411
24. ^Microsoft Learn. "Mistral AI / Mistral Large retirement on Azure AI Studio." April 2025. learn.microsoft.com/...2150001
25. ^1 ^2 ^3Mistral AI. "Introducing Mistral Large 4." October 6, 2026. mistral.ai/...mistral-large-4
26. ^1 ^2 ^3Mistral AI. "Mistral Large 4 (Le Chonk) is now open weight." October 29, 2026. mistral.ai/...mistral-large-4-oss
27. ^Mistral AI Help Center. "Le Chat is now Vibe." Accessed October 30, 2026. help.mistral.ai/...682992-le-chat-is-now-vibe
28. ^TechCrunch. "Mistral AI releases new model to rival GPT-4 and its own chat assistant." February 26, 2024. techcrunch.com/...gpt-4-and-its-own-chat-assistant
29. ^Mistral AI. "AI in abundance." September 17, 2024. mistral.ai/...september-24-release
30. ^Mistral AI. "Pixtral Large." November 18, 2024. mistral.ai/...pixtral-large