by Mistral AI

Mistral Large 4 review, pricing and verdict

Mistral AI's largest model to date, a reasoning-capable multimodal MoE for coding, agents and cybersecurity, with public weights under a modified MIT license.

Best for
Security research and vulnerability work
Not for
Reselling inference as an API
  • Available
  • Open weights

Should you use Mistral Large 4?

Mistral Large 4 left preview in early November 2026 with downloadable weights under a modified MIT license that only restricts resale as a hosted service. It scores 26 on Artificial Analysis Intelligence Index v5.0.1, above the 23 median for open-weight models its size, and serves a 1M-token context.

Best for

  • Security research and vulnerability work Mistral reports its strongest results on cyber tasks, and self-hosted weights avoid provider refusals mid-incident.
  • EU-sovereign or on-premise deployments Mistral serves it from its own European datacenters, and the public weights can run on your own hardware.
  • Agents that read images and call tools One model handles image input, function calling and strict JSON schemas, with three reasoning effort levels.

Not for

  • Reselling inference as an API The license requires a separate commercial agreement with Mistral for Model as a Service businesses.
  • Cheapest general intelligence per dollar Artificial Analysis scores it 26 on its v5.0.1 index, behind the leading Chinese open-weight models on that aggregate.
  • Single-workstation self-hosting The full 1.05T weights need a multi-GPU server even though only about 5% of parameters fire per token.

Can you use Mistral Large 4?

  • Free plan: not published.
  • Open weights: released (Mistral Ecosystem Open-Weight License (modified MIT; Model-as-a-Service businesses need a commercial license)).

Mistral Large 4 pricing

Input, per 1M tokens
$0.70/1M
Output, per 1M tokens
$2.10/1M
Cached input, per 1M tokens
$0.07/1M
Pricing model
per-token
Free tier
No

What a real job costs

JobInputOutputTotal
Summarise a 20-page PDF$0.021$0.0021$0.023
Support reply$0.0014$0.0006$0.0020
One coding agent run$0.140$0.042$0.182

Budgets: 20-page PDF = 30k in / 1k out · Support reply = 2k in / 300 out · Coding agent run = 200k in / 20k out. Computed from the vendor's per-token prices at render time; cached-input discounts are not applied.

Mistral lists the API at $0.70 per 1M input tokens, $0.07 per 1M cached input tokens and $2.10 per 1M output tokens, read on its pricing page on 9 November 2026. During the October preview the page showed a 50% sale price of $0.68 input and $2.09 output against crossed-out list prices of $1.36 and $4.18; that sale label no longer appears. Mistral Large 3 is listed at $0.50 input and $1.50 output per 1M tokens. Batch and Priority rates sit on separate tabs of Mistral's pricing page and are not recorded here. The open weights can also be self-hosted with no per-token fee.

Self-hosting requirements

Self-hosted
Yes
Recommended GPU
Multi-GPU server; Mistral's examples use an 8-GPU node with tensor parallelism, and the NVFP4 checkpoint targets NVIDIA Blackwell GPUs

What Mistral Large 4 can do

Mistral Large 4 is Mistral AI's 1.05 trillion parameter open-weight Mixture-of-Experts model, generally available since November 2026 after an October preview. About 49 billion parameters fire per token. It takes text and images, serves a 1M-token context, and costs $0.70 per million input tokens on Mistral's API, read 9 November 2026.

What goes in and what comes out

Reads

  • text
  • image

Writes

  • text
  • tool-calls

Capabilities

  • Function callingSupported
  • Structured outputSupported
  • Tool useSupported
  • VisionSupported
  • Audio inputNot supported
  • Audio outputNot supported
  • Video inputNot supported
Reasoning modes
low, medium, high

Key features

  • Granular Mixture-of-Experts

    Sparse routing picks a handful of experts per token, so roughly 5% of the weights do the work on each step.

  • 1M-token context

    Prompts up to 1,048,576 tokens and replies up to 256K tokens, per the model card and OpenRouter.

  • Three reasoning levels

    Low, medium and high reasoning effort per request, with high as the default.

  • Agent and tool support

    Function calling, strict JSON schemas, document question answering, batching, and Agents and Conversations endpoints with built-in tools.

  • Open weights with NVFP4

    Weights on Hugging Face plus an official NVFP4 checkpoint for Blackwell GPUs, served through vLLM or SGLang.

Known limitations

  • Selling hosted API access to third parties needs a separate commercial license from Mistral, unlike Large 3's Apache 2.0.
  • Self-hosting the full weights needs a multi-GPU server; Mistral's examples assume an 8-GPU node.

Who it is for

  • Security teams that need an open-weight model for vulnerability research and incident response
  • Teams building document and image analysis pipelines
  • Engineering groups running agentic coding and business-workflow agents
  • European organizations that want a vendor-operated EU deployment under European law

Mistral Large 4 benchmarks

Mistral Large 4 against its nearest rivals

Output speed tokens per second, higher is better

Standout result

93%

Cybench challenges solved (vendor-reported)

Source: Mistral Large 4 launch post, 6 October 2026

Where it sits

Priced around the middle of the 106 GA models with a published price (rank 45), rank 24 of 59 on output speed as cited from Artificial Analysis, and one of 41 that document a zero-data-retention option.

MeasureThis modelRankMedian
Blended price (3:1)$ per 1M tokens$1.05/M#45 / 106peer median $1.73/M
Output speedtokens/s112 tok/s#24 / 59peer median 94 tok/s

Ranks are against GA models on Hokai that publish the same figure; ties share a rank.

Benchmark scores

  • AA Intelligence Index cited: Artificial Analysis · 09 Nov 202626Composite of the evaluations Artificial Analysis runs for its index, 0 to 100.

Measured figures

Output speed
112 tok/s

A benchmark is an exam, not the job. Scores transfer unevenly between tasks, so weigh the one closest to your workload and read every figure with its source.

Safety benchmarks

Source
vendor-reported, Mistral Large 4 launch post
Kora score of 2
1.691
B3 attack resistance pct
93.3

Strengths and limits

Strengths

  • Strong vendor-reported cyber results from a model security teams can now run on their own hardware.
  • Image input, tool calls and strict JSON schemas in one open-weight model with a 1M-token context.
  • Mistral reports 93.3% resistance on Lakera's B3 agent security benchmark.
  • Mistral reports 82% on CyberGym-E2E, a test that asks a model to reproduce a real vulnerability and then patch it.
Show 2 more
  • Public weights under a modified MIT license let security and EU teams self-host it, with only resale as a hosted service restricted.
  • Mistral's model card cites 48% on the Artificial Analysis Coding Agent Index while using 47% fewer tokens per task than GLM-5.3.

Limits

  • Hosting providers need a separate commercial license, so the terms are narrower than Large 3's Apache 2.0.
  • It trails the leading Chinese open-weight models on the Artificial Analysis v5.0.1 aggregate.
  • Defaults to high reasoning effort, which raises token spend unless you lower it.

Mistral Large 4 alternatives

The nearest AI models on Hokai for the same job.

ModelLabContextIn / out per 1MReads and writes
Mistral Large 4 This pageMistral AI1M$0.70 / $2.10text, image → text
Claude Opus 5.5Anthropic's September 2026 flagship LLM with a 1 million token context window, built for long-running agentic coding and computer use.Anthropic1M$4 / $20text, image, pdf → text
Gemini 4 ArgonGoogle DeepMind's efficient Gemini 4 reasoning model, generally available on the paid Gemini API: 1M-token context and output, text, image, video, audio and PDF input.Google1M$2 / $10text, image, video, audio → text
Hy4 flashTencent's Hy4 flash, released 16 November 2026, is a lower-cost multimodal model that reads text and images across a 1M-token window.Tencent1M$0.07 / $0.281text, image → text

Mistral Large 4 specs

Context window
1.0M tokens
Max output
262K tokens
Tokens per second
112 tok/s
Architecture
Mixture-of-Experts (granular)
Mixture of experts
Yes
Parameters
1.05T total / 52B active (MoE), plus 1.6B vision encoder

Training data

Trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's European datacenters, with a significant share of multilingual data across more than 160 languages, then post-trained with supervised fine-tuning and large-scale reinforcement learning. Mistral has not published a knowledge cutoff.

API endpoints

NameEndpointAuth
Mistral Studio API (model id mistral-large-4)https://docs.mistral.ai/models/mistral-large-4-0API key
OpenRouter (Mistral as sole provider)https://openrouter.ai/mistralai/mistral-large-4-0OpenRouter API key
Hugging Face weights (self-host with vLLM or SGLang)https://huggingface.co/mistralai/Mistral-Large-4-1T-A52BNone (public weights)

Prompting patterns

Show 2 prompting patterns
  • Match reasoning effort to the task

    Use reasoning_effort "low" for chat replies and extraction, "medium" for day-to-day work and "high" for agent runs.

    Mistral's model card recommends this split and makes high the default, which spends the most tokens.

  • Cap output length explicitly

    Set max_tokens well below the 256K ceiling for chat and tool-call turns.

    Artificial Analysis rates the model somewhat verbose, so uncapped completions raise per-task cost.

Known quirks

Show 4 known quirks
  • Two release dates in circulation

    Artificial Analysis dates the release 29 October 2026, while Mistral's model card marks version 26.10 GA on 5 November 2026; the hosted preview opened on 6 October.

    Workaround: Cite the date for the event you mean: preview, weights or GA.

  • High reasoning is the default

    Reasoning effort defaults to high, the most token-hungry setting, and Artificial Analysis counted 107 million output tokens on its index.

    Workaround: Set reasoning_effort to low or medium for routine calls and cap max_tokens.

  • Open weights, restricted resale

    The weights are downloadable under a modified MIT license, but a Model as a Service business needs a separate commercial license.

    Workaround: Check whether your product gives third parties meaningful control over the model before you ship it as an API.

  • Parameter count rounds differently

    Mistral's docs list 1.05T total and 52B active, while the Hugging Face card says about 1T total and 49B active excluding embeddings and output layers.

    Workaround: Size hardware from the checkpoint files, not the headline number.

Mistral Large 4 facts

Deployment
API · OpenRouter · Self-hosted
Safety posture
Balanced
Released
06 Oct 2026
Current version
mistral-large-4-0
Family
Mistral Large
Variant
Multimodal
Licence
Mistral Ecosystem Open-Weight License (modified MIT) · Open weights
Data retention
Mistral's help center says Studio free mode may use input and output data for training with an opt-out, pay-as-you-go customers can opt out at any time, and Vibe Enterprise is opted out by default. A Zero Data Retention option is documented in the help center. A retention period specific to this model is not stated.

Content restrictions

  • Mistral reports a refusal rate on harmful cyber prompts (JailbreakBench, StrongREJECT, AgentHarm) higher than any open-weight model it compared.
  • Mistral says vetted security partners and state authorities red-team a version with reduced moderation before weights ship.

Safety controls

Red teaming
Vendor-described: cybersecurity leaders, vetted partners and state authorities, ahead of the weights release

Data governance

Source URL
https://help.mistral.ai/en/articles/347617-do-you-use-my-user-data-to-train-your-artificial-intelligence-models
Zero data retention option documented
Yes

Version history

  • mistral-large-4-0 (v26.10, GA) · 05 Nov 2026

    Generally available with public weights on Hugging Face under the Mistral Ecosystem Open-Weight License, 1M context, NVFP4 checkpoint.

  • mistral-large-4-0 (v26.10, preview) · 06 Oct 2026

    Hosted public preview: 1.05T total and 52B active parameters, image input, sale-priced API.

  • mistral-large-3 (v25.12) · 02 Dec 2025

    Predecessor: 675B total and 41B active parameters, listed by Mistral under Apache 2.0.

What changed

  • Release Mistral Large 4: Generally available, weights public under the Mistral Ecosystem Open-Weight License (mistral-large-4-0 (v26.10)) source
  • Benchmark Artificial Analysis scores the preview at 38 on its Intelligence Index (mistral-large-4-0) source
  • Release Mistral Large 4 launches in public preview (mistral-large-4-0) source

Questions about Mistral Large 4

How much does the Mistral Large 4 API cost per million tokens?

Mistral lists $0.70 per million input tokens, $0.07 for cached input and $2.10 per million output tokens on its pricing page, read on 9 November 2026. The October preview carried a 50% "sale" label against crossed-out prices twice as high; that label is gone and the listed rate is now the only one shown. OpenRouter passes the same rates through with Mistral as the sole provider.

Is Mistral Large 4 better than DeepSeek V4 Pro 0813?

On Mistral's own charts it leads DeepSeek V4 Pro 0813 on its Coding Agent Index and on AutomationBench, but those are vendor results. On the independent Artificial Analysis aggregate, Mistral's own model card concedes that the leading Chinese open-weight models score higher. Pick Large 4 for image input, cyber work and EU hosting; pick DeepSeek for general reasoning per dollar under its MIT weights.

Can I use Mistral Large 4 weights commercially?

Yes, in most cases. The Mistral Ecosystem Open-Weight License is a modified MIT license that allows commercial use, fine-tuning and redistribution. The exception is a Model as a Service business that sells third parties API or hosted access to the model: it needs a separate commercial license from Mistral. Internal use and products that embed the model in specific features are exempt.

Will Mistral train on what I send to Large 4?

Mistral's help center says Studio free mode may use inputs and outputs for training with an opt-out, pay-as-you-go customers can opt out at any time, and Vibe Enterprise is opted out by default. Thumbs-up or thumbs-down feedback lets Mistral use that exchange. A Zero Data Retention option is documented, and self-hosting the open weights keeps data off Mistral's servers entirely.

Who is Mistral Large 4 a good fit for, and who should pass?

It fits security teams, EU organizations that want a European-run or self-hosted model, and agent builders who need image input with tool calls. Pass if you only want the most capable model per dollar, if you cannot staff an 8-GPU deployment and do not want the API, or if you resell inference and do not want to negotiate a commercial license.

About Mistral Large 4

Mistral Large 4 is the largest model in Mistral AI's Large line and, since early November 2026, a generally available open-weight release rather than a preview. Mistral first opened it on 6 October 2026 as a hosted preview nicknamed "le Chonk". Artificial Analysis dates the weights-backed release 29 October 2026, and Mistral's model card now marks version 26.10 as GA, dated 5 November 2026. It is a granular Mixture-of-Experts model with 1.05 trillion total parameters per the model card (the Hugging Face card rounds this to about 1 trillion), 49 billion active per token or 52 billion counting embeddings and output layers, 128 routed experts with 4 active plus 1 shared, and a 1.6 billion parameter vision encoder. Mistral says it trained the model from scratch on 3,800 NVIDIA Grace Blackwell GPUs in its own European datacenters. It follows Mistral Large 3, the 675B total, 41B active model from December 2025, and is built by Mistral AI.

Read the full background

The weights are public on Hugging Face as Mistral-Large-4-1T-A52B under the Mistral Ecosystem Open-Weight License, which Mistral describes as a modified MIT license. It allows commercial and non-commercial use, modification, fine-tuning and redistribution, with one condition: a company that runs a Model as a Service business, meaning it gives third parties API or hosted access with meaningful control over inputs, parameters or training data, must sign a separate commercial license with Mistral first. Internal use, end-user products that embed the model inside specific features, and use through Mistral's own products are exempt. That makes it more restrictive than Large 3's Apache 2.0 for hosting providers, and close to MIT for everyone else. Artificial Analysis files it under open weights with commercial use restricted.

Mistral publishes most results as charts, so every number in this paragraph is vendor-reported unless it names another source. On agentic coding Mistral reports 61.7% on DeepSWE v1.1, 59.4% on SWE-Atlas-QnA and 28.3% on Terminal-Bench 4, and places it ahead of DeepSeek V4 Pro 0813 and Qwen3.8 Max on its own Coding Agent Index. Its Hugging Face card cites a 48% score on the Artificial Analysis Coding Agent Index, which it calls effectively tied with GLM-5.3 while using 47% fewer tokens per task. In cybersecurity it reports 93% on Cybench and 82% on CyberGym-E2E, a test that asks a model to reproduce a real vulnerability and then patch it. A blind Surge AI rating of coding output put it second of five at 3.74 out of 5, behind Claude Opus 5 and ahead of Kimi K3.

The independent view is more modest. Artificial Analysis now scores it 26 on Intelligence Index v5.0.1, a new nine-evaluation version of the index, against a median of 23 for open-weight models of its size, so scores from the older v4 index (where the preview scored 38) do not compare. It measured 112 output tokens per second and 107 million output tokens to finish the index, against a median of 99.3 million, and it describes the model as somewhat verbose. Running the index cost $2.06 per task at Mistral's rates. The model leaderboard sets those scores next to current prices.

The context window is 1M tokens on the model card, on Artificial Analysis and on OpenRouter, which also lists a 256K output cap, so the 524K figure the preview endpoint served in October no longer applies. Input is text and images and output is text. The API supports function calling, structured outputs with strict JSON schemas, document question answering, prefix completion, batching, and the Agents and Conversations endpoints with built-in tools. Reasoning effort has three settings, low, medium and high, with high as the default. Mistral says the training data covers more than 160 languages, including every official EU language.

Self-hosting is now possible but heavy. Mistral recommends vLLM and ships an official Docker image, its examples assume an 8-GPU node with tensor parallelism, and a quantized NVFP4 checkpoint targets Blackwell GPUs with a smaller memory footprint. Only about 5% of the parameters fire per token, which keeps compute down, but the full weights still have to sit in GPU memory, which rules out a single workstation. The local AI guide covers what smaller setups can run instead.

On safety, Mistral reports 93.3% attack resistance on Lakera's B3 agent security benchmark and 1.691 of a possible 2 on the KORA benchmark, and says its refusal rate on harmful cyber prompts is higher than any open-weight model it compared. It also says Claude Opus 5.5 and GPT-6 Astra score near zero on one cyber test because they refuse it, which is Mistral's claim, not an independent finding.

Who should not pick it: a team that only wants the cheapest capable API gets more general-index score per dollar from smaller models, and the leading Chinese open-weight models score higher on the Artificial Analysis aggregate, which Mistral's own card concedes. A hosting provider that plans to resell inference needs Mistral's commercial license before launch, and a team that wants plain Apache 2.0 terms can still run Mistral Large 3. Hokai's verdict on what Large 4 is for walks through the jobs where it does fit: security work, EU-sovereign deployments and self-hosted agents.

Sources and guides

More AI Models on Hokai

  • Nemotron-Labs-SigmaNVIDIA's Nemotron-Labs-Sigma, released in November 2026 in 8B and 3B sizes, writes text by denoising whole blocks of tokens in parallel, not one at a time.
  • Qwen4 MaxAlibaba's Qwen4 Max, released 16 November 2026, is a sparse mixture-of-experts model with a 1M-token context and text, image and video input.
  • Gemini 4 HydrogenGoogle DeepMind's November 2026 Gemini 4 model for coding agents, with four thinking levels from low to xhigh and a free tier in AI Studio.
  • Qwen4-Max-PreviewAlibaba's November 2026 Qwen 4 flagship preview, open only to Token Plan subscribers, with three reasoning levels from low to xhigh.
  • Claude Fable 5Superseded by Fable 5.1 and Fable 5.2 but still sold on the Claude API: Anthropic's first generally available Mythos-class model, released June 2026, with a 1M-token context.
  • Claude Fable 5.1Superseded by Claude Fable 5.2 in November 2026 but still sold on the Claude API: Anthropic's second Mythos-class model, released Sept 1, 2026, for long-horizon coding and research.