Skip to content
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed
  1. Home/
  2. Compare/
  3. Mistral Large 2 vs Mistral Small 4
Updated Mar 2026

Mistral Large 2 vs Mistral Small 4

Mistral Small 4 is cheaper than Mistral Large 2 at $0.15/M vs $8/M input tokens.

Pokémon Battle UI Recreation

Pokémon battle screen you can actually click. One HTML file.

Loading the build
Mistral Large 2
Loading the build
Mistral Small 4

Which answer wins?

Written rounds

11 prompts
  • Three minutes of stand-up. Puns are banned.

    Mistral Large 2: Alright, good evening everybody! How are we all doing tonight? Yeah? Yeah? nobody's gonna say 'bad'? Ok, fair enough. You know, I was walking down the street the other day, and I saw this sign outside a shop that said, "Wet Paint." I thought, "Well, that's weird. Isn't paint supposed to be wet?

    Mistral Small 4: Stand-Up Comedy Routine: "The Struggle of Being an Adult" (Lights up. Comedian walks on stage, mic in hand, casual vibe.) Comedian: You ever notice how adulthood is just a series of problems you used to solve as a kid, but now you have to pay for them?

    Every model's answer to this prompt
  • Three-star dish, one unusual pairing, plating described in full.

    Mistral Large 2: Michelin 3-Star Dish: "Ocean’s Whisper & Forest’s Breath" Conceptual Narrative: This dish is a poetic dialogue between two seemingly disparate ecosystems—the ocean’s depths and the forest floor—united through texture, temperature, and memory.

    Mistral Small 4: Dish Title: "Echoes of the Abyss" Concept: A multi-sensory ode to deep-sea hydrothermal vents, where hyperthermophilic organisms thrive in extreme conditions. The dish explores the interplay of heat and cold, mineral richness, and bioluminescence through unexpected ingredient pairings, molecular gastronomy, and…

    Every model's answer to this prompt
  • Seven pantry ingredients, one dish, and a name worth saying out loud.

    Mistral Large 2: Here’s a simple, delicious, and creative savory dish using just 7 common pantry ingredients—no fresh produce required! "Golden Crispy Chickpea & Tomato Skillet" A warm, spiced, and satisfying one-pan dish with crispy chickpeas, tangy tomatoes, and a hint of smoky sweetness. Perfect as a side, snack, or light meal!

    Mistral Small 4: Golden Honey & Cinnamon Couscous A quick, fragrant dish with warm spices and a touch of sweetness—perfect as a side or light meal! Ingredients (7): 1 cup couscous (uncooked) 1 ¼ cups water or broth 2 tbsp honey 2 tbsp butter or olive oil ½ tsp ground cinnamon ¼ cup raisins or chopped nuts (optional but recommended)…

    Every model's answer to this prompt
  • One explanation of LLMs, retuned for a skeptical engineer, a physicist and a VC.

    Mistral Large 2: 1. For the Experienced Software Engineer (Skeptical, Systems-Minded, Non-ML Background) You’re right to be skeptical—"predicting the next word" sounds trivial, like autocomplete on steroids. But the magic isn’t in the prediction itself; it’s in the scale and architecture of how that prediction is trained.

    Mistral Small 4: 1. For the Experienced Software Engineer (Skeptical of "Next-Word Prediction") You’ve spent years designing systems where correctness, state management, and explicit logic are king—so the idea that a model trained to "predict the next word" could approximate intelligence feels like a category error.

    Every model's answer to this prompt
  • You are the CEO. The drug has a side effect nobody reported. The board meets in 48 hours.

    Mistral Large 2: As CEO, your immediate priority must be patient safety while balancing legal, ethical, regulatory, and financial obligations. Delaying disclosure is unethical, legally risky (potential criminal liability under the False Claims Act or FDCA), and could irreparably damage trust.

    Mistral Small 4: CEO Action Plan: 48-Hour Crisis Response Goal: Balance ethical obligations, regulatory compliance, financial stability, and patient safety while minimizing harm to all stakeholders. Hour 0-6: Immediate Assessment & Core Team Assembly 1.

    Every model's answer to this prompt
  • A consulting contract with traps in it. Find them, then fix them.

    Mistral Large 2: This contract is heavily one-sided in favor of the Client (TechCorp Solutions Inc.) and contains multiple clauses that could be exploited, abused, or create unreasonable legal and financial risks for the Contractor.

    Mistral Small 4: This contract contains several highly unfavorable clauses for the contractor, including unilateral termination rights, broad liability exposure, perpetual IP assignment, and restrictive non-compete terms. Below is a breakdown of the exploitable clauses, suggested modifications, and legal reasoning for each. 1.

    Every model's answer to this prompt
  • The transistor arrives in 1920. Trace it forward to 1980, second-order effects included.

    Mistral Large 2: The invention of the transistor in 1920—nearly three decades earlier than its actual debut in 1947—would have triggered a cascade of technological, economic, and geopolitical shifts with profound second- and third-order effects. Below is a detailed exploration of these implications through 1980, structured by domain.

    Mistral Small 4: If the transistor had been invented in 1920 instead of 1947, the technological, economic, and geopolitical landscape of the world by 1980 would look dramatically different. Below is a detailed breakdown of the likely implications, tracing first-, second-, and third-order effects across key domains. 1.

    Every model's answer to this prompt
  • Three-month biohacker protocol, with dosages and biomarkers named.

    Mistral Large 2: Ultimate 3-Month Longevity & Biohacking Optimization Plan Designed for a high-performance biohacker seeking to maximize lifespan, cognitive function, physical performance, and metabolic resilience.

    Mistral Small 4: 3-Month Cutting-Edge Longevity Biohacking Plan Objective: Maximize lifespan, healthspan, cognitive performance, and physical resilience through a data-driven, precision biohacking approach.

    Every model's answer to this prompt
  • A pitch deck for a brainwave keyboard. The 94% accuracy claim is doing a lot.

    Mistral Large 2: Here are the three weakest claims in MindMeld AI’s pitch deck, along with explanations for why they’re problematic and concrete suggestions to strengthen them: 1.

    Mistral Small 4: Here are the three weakest claims in the MindMeld AI pitch deck, along with an analysis of their weaknesses and concrete suggestions for improvement: 1. Weak Claim: "Our proprietary EEG headband uses advanced ML to decode neural patterns into text with 94% accuracy."

    Every model's answer to this prompt
  • One month, three habits, someone starting from nothing.

    Mistral Large 2: Here’s a simple, actionable 1-month plan to kickstart your journey toward better health and longevity. We’ll focus on three key pillars: nutrition, movement, and recovery (sleep + stress management). Each week builds on the last, with small, sustainable changes.

    Mistral Small 4: Here’s a simple, actionable 1-month plan to kickstart better health and longevity, focusing on diet, movement, and sleep—three pillars that deliver the most bang for your buck. Each week builds on the last, with small, sustainable changes.

    Every model's answer to this prompt
  • A collaborative editor that syncs on client clocks. Find what breaks first.

    Mistral Large 2: Here’s a comprehensive analysis of potential failure modes, race conditions, and scaling bottlenecks in this architecture, along with proposed solutions and their trade-offs: 1.

    Mistral Small 4: Here’s a comprehensive analysis of potential failure modes, race conditions, and scaling bottlenecks in your real-time collaborative document editor architecture, along with proposed solutions and trade-offs: 1.

    Every model's answer to this prompt

Favorites

Movie

Album

Same pick

Book

City

Same pick

Game

Mistral Large 2Mistral Large 2

The Shawshank Redemption

1994

OK Computer

Radiohead

La sombra del viento

Carlos Ruiz Zafón

Kyoto

Japan

The Legend of Zelda: Ocarina of Time

Action

Mistral Small 4Mistral Small 4

The Dark Knight

2008

OK Computer

Radiohead

Братья Карамазовы

Fiódor Dostoievski

Kyoto

Japan

The Legend of Zelda: Breath of the Wild

Adventure, Action

Price and specs

Mistral Large 2 and Mistral Small 4 compared across 23 shared prompts
SpecMistral Large 2Mistral Small 4
Input price$8/M tokens$0.15/M tokens
Output price$24/M tokens$0.6/M tokens
Context window128K tokens262K tokens
WeightsOpenOpen
Free API (OpenRouter)NoNo
ReleasedJul 2024Mar 2026
At 10M a month$80.00$80.00$1.50$1.50
1M10M100M1B10M tokens

Input tokens at list price. No caching, no batch discount.

Where to run it2 hosts
Mistral Large 21 host
HostInOutContextUptime
  • Mistral$2.00 in·$6.00 out·131k·50% up
Mistral Small 41 host
HostInOutContextUptime
  • Mistral$0.15 in·$0.60 out·262k·99.1% up

Per million tokens. Prices and uptime via OpenRouter, checked 10 Oct 2026.

Common questions

What is the difference between Mistral Large 2 and Mistral Small 4?

Both are developed by Mistral AI but target different use cases. Mistral Large 2 has a 128K token context window vs Mistral Small 4's 262K. You can compare their actual outputs across 23 challenges on Rival to see how they differ in practice.

Which is better, Mistral Large 2 or Mistral Small 4?

It depends on your use case. Mistral Large 2 and Mistral Small 4 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 23 challenges so you can judge which fits your needs best.

How much does Mistral Large 2 cost compared to Mistral Small 4?

Mistral Large 2 costs $8/M input tokens and Mistral Small 4 costs $0.15/M input tokens. Mistral Small 4 is $7.85/M cheaper per input. Check their side-by-side outputs on Rival to see if the price difference is justified by quality.

How can I compare Mistral Large 2 and Mistral Small 4 on Rival?

This page shows a side-by-side comparison of Mistral Large 2 and Mistral Small 4 across shared challenges. You can vote on which model produced the better output in a blind duel. Browsing and voting are free. No account is needed to look; signing in only saves your votes and likes.

More comparisons

Against the newest arrivals

  • Mistral Large 2 vs Step 5 PreviewLanded Oct 2026
  • Mistral Small 4 vs Claude Haiku 5.5Landed Oct 2026
  • Mistral Large 2 vs Ling 3.1 FlashLanded Oct 2026
  • Mistral Small 4 vs Mistral Large 4Landed Oct 2026
  • Mistral Large 2 vs GPT-6.1 SolLanded Sep 2026
  • Mistral Small 4 vs Claude Sonnet 5.5Landed Sep 2026
  • Mistral Large 2 vs Solar Mini 4Landed Sep 2026
  • Mistral Small 4 vs Qwen3.8 Max PrimeLanded Sep 2026

Same lab, same size, long tail

  • Mistral Large 2 vs Mistral Large 4Version compare
  • Mistral Large 2 vs Mistral Large 3 2512Same lab
  • Mistral Small 4 vs Mistral Large 3 2512Same lab
  • Mistral Small 4 vs Mistral Medium 3.1Same lab
  • Mistral Small 4 vs GLM 5New provider
  • Mistral Large 2 vs GLM 5 TurboNew provider
  • Mistral Large 2 vs GLM 5.1Same size
  • Mistral Large 2 vs GLM 5.2New provider

Model pages

  • Mistral Large 223 outputs, specs and price
  • Mistral Small 453 outputs, specs and price
Workshop · 150 skills

Build your own version.

Editable skills and working examples for your agent.

Explore $79Free sample
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed