What is Mistral Large 3?

Mistral Large 3 is an advanced AI model suited for demanding production workloads, developed by Mistral and released in 2025-12. It handles general AI tasks including writing, analysis, summarisation, and question answering.

Is Mistral Large 3 open-source or open-weight?

Mistral Large 3 is an open-weight model. Its trained weights are publicly available for download, self-hosting, and fine-tuning. You can run it on your own infrastructure without per-token API fees. Note that open-weight is not the same as open-source: the training code and data may not be fully public.

What benchmarks has Mistral Large 3 been evaluated on?

Mistral Large 3 has been evaluated on: GPQA Diamond (43.9%), MMLU-Pro (73.1%), MATH-500 (93.6%). Scores tagged SR are self-reported by the provider; scores tagged 3P or AA have been independently verified by third parties.

Is Mistral Large 3 available in FlowHunt?

Yes — Mistral Large 3 is directly available in FlowHunt. You can use it to build AI agents, automated workflows, chatbots, and document processing pipelines without any additional setup.

How does Mistral Large 3 compare to other AI models?

You can compare Mistral Large 3 against all 24 models tracked on the FlowHunt AI Leaderboard using the sortable benchmark table. The Benchmark Rankings section on this page shows Mistral Large 3's position on each benchmark it has been evaluated on.

Mistral

Mistral Large 3

AdvancedOpen-weightAvailable in FlowHunt

Try in FlowHunt

Mistral Large 3 is an advanced AI model suited for demanding production workloads, developed by Mistral and released in December 2025. As an open-weight model, its trained weights are publicly available — you can self-host it, fine-tune it on proprietary data, or run it on-premise without API dependencies. It supports a 256K-token context window, adequate for most documents and multi-turn applications.

On the FlowHunt AI Leaderboard it is one of 24 tracked models, evaluated across 3 benchmarks. Standout scores: 93.6% on MATH-500 (#1 of 4); 73.1% on MMLU-Pro (#5 of 8); 43.9% on GPQA Diamond (#14 of 14).

Mistral Large 3 was released on December 2, 2025 as Mistral’s most capable commercial model and, critically, as an open-weight release under the Apache 2.0 license. At 41B active parameters from a 675B total MoE architecture, it is the largest and most capable fully Apache-licensed frontier-adjacent model available. Mistral designed Large 3 to address European enterprise needs specifically: deployable in EU data centres, compliant with GDPR by architecture (data never leaves the deployment region), and usable for commercial applications without the restrictive licensing terms of many open-weight models. Its MMLU-Pro score of 84.2% and solid science reasoning profile have made it the default choice for European regulated industries — finance, healthcare, and legal — that need strong AI capability without sending data to US-based API providers.

Note: Enterprise generalist, Apache 2.0. Weak hard reasoning (GPQA 43.9%).

Try in FlowHunt

Released: December 2025
Parameters: 41B/675B MoE
Context: 256K tokens
Weights: Open
Tier: Advanced
Provider: Mistral

Mistral Large 3 Benchmark Rankings

Rank among all 24 models tracked on this leaderboard. SR = self-reported · 3P / AA / C = independently verified.

Benchmark	Score	Rank	Source	What it measures
GPQA Diamond	43.9%	14 of 14	3P	Graduate-level Google-proof science Q&A
MMLU-Pro	73.1%	5 of 8	3P	Hard knowledge reasoning, 12K questions
MATH-500	93.6%	1 of 4	3P	Hendrycks math (500 problems)

Detailed Benchmark Scores

GPQA ◇ 3P

43.9%

Graduate-level Google-proof science Q&A

MMLU-P 3P

73.1%

Hard knowledge reasoning, 12K questions

MATH 3P

93.6%

Hendrycks math (500 problems)

Mistral Large 3 vs. Advanced Peers

Head-to-head benchmark comparison with other Advanced-tier models. Higher is better for all metrics.

Benchmark	Mistral Large 3	Claude Sonnet 4.6	GPT-4.1
GPQA ◇	43.9%	—	—
MMLU-P	73.1%	—	—
MATH	93.6%	89.0%	—

About Mistral

Mistral Founded 2023 · Paris, France

Website →

Mistral AI is a French AI company founded in April 2023 by Arthur Mensch (formerly of Google DeepMind), Guillaume Lample, and Timothée Lacroix (both formerly of Meta AI Research). The company positions itself as a European champion in generative AI, combining a commercial API business with strategic open-weight releases. Mistral's flagship Large series and Apache 2.0-licensed models have made it the most prominent non-US, non-Chinese provider on most AI leaderboards. The company maintains EU headquarters in Paris, emphasises GDPR and EU regulatory compliance, and has articulated a distinct regulatory philosophy calling for lighter regulation of open-weight models compared to closed API systems.

Use Cases

What to Use Mistral Large 3 For

Scientific & Technical Reasoning: Mistral Large 3 achieves 43% on GPQA Diamond (#14 of 14 models). It handles general scientific Q&A at a solid level, though frontier models score 25+ percentage points higher on the hardest graduate-level problems.
Self-Hosted & Private Deployments: Mistral Large 3 is open-weight (41B/675B MoE parameters) — its trained weights are publicly available for download and self-deployment. Run it on your own GPU hardware or private cloud to ensure data never leaves your environment, eliminate per-token API costs at scale, or fine-tune the model on proprietary datasets.

Strengths & Limitations

Strengths

Advanced mathematics — 93% MATH-500
Self-hosted & data-private deployments (open-weight)
Long-context tasks (256K context window)

Limitations

Below-frontier hard science reasoning (43% GPQA — frontier is 90%+)

More to Compare

Related AI Models

Browse the Full AI Model Leaderboard

Compare all 24 models across 11 benchmarks — sortable, sourced, and updated June 2026.

← Back to AI Leaderboard Try in FlowHunt

Frequently asked questions

: Mistral Large 3 is an advanced AI model suited for demanding production workloads, developed by Mistral and released in 2025-12. It handles general AI tasks including writing, analysis, summarisation, and question answering.
: Mistral Large 3 is an open-weight model. Its trained weights are publicly available for download, self-hosting, and fine-tuning. You can run it on your own infrastructure without per-token API fees. Note that open-weight is not the same as open-source: the training code and data may not be fully public.
: Mistral Large 3 supports a 256K-token context window, adequate for most documents and multi-turn applications.
: Mistral Large 3 has been evaluated on: GPQA Diamond (43.9%), MMLU-Pro (73.1%), MATH-500 (93.6%). Scores tagged SR are self-reported by the provider; scores tagged 3P or AA have been independently verified by third parties.
: Yes — Mistral Large 3 is directly available in FlowHunt. You can use it to build AI agents, automated workflows, chatbots, and document processing pipelines without any additional setup.
: You can compare Mistral Large 3 against all 24 models tracked on the FlowHunt AI Leaderboard using the sortable benchmark table. The Benchmark Rankings section on this page shows Mistral Large 3's position on each benchmark it has been evaluated on.

Learn more

Mistral Medium 3 — Benchmarks, Capabilities, and Use Cases

Mistral Medium 3 is a proprietary AI model from Mistral, classified as Advanced tier with a 128K tokens context window. Explore benchmark scores, strengths, and...

May 1, 2025 2 min read

MiniMax M3 — Benchmarks, Capabilities, and Use Cases

MiniMax M3 is a open-weight AI model from MiniMax (MoE (undisclosed)), classified as Frontier tier with a 1M tokens context window. It is available directly in ...

Jun 1, 2026 4 min read

Mixtral 8x7B — Benchmarks, Capabilities, and Use Cases

Mixtral 8x7B is a open-weight AI model from Mistral (12.9B/46.7B MoE), classified as Legacy tier with a 32K tokens context window. It is available directly in F...

Dec 1, 2023 3 min read