Compare / MiniMax M3.1 Flash vs Mistral Large 4
MiniMax M3.1 Flash vs Mistral Large 4
Which is cheaper, which is more capable, and which does better at each kind of work, from 82 results on 5 sources that measured both models.
MiniMaxMiniMax M3.1 Flash
Mistral AIMistral Large 4
MiniMax M3.1 Flash wins 3410 too close to callMistral Large 4 wins 38
Of the 82 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 20 October 2026 (catalogue release 2026-10-20-96eb40c3ec63). See also MiniMax vs Mistral AI, every model from both labs.
MiniMax M3.1 Flash or Mistral Large 4? The short answer
Pick MiniMax M3.1 Flash: it scores higher (Spring Prompt overall 47 against 33) and costs less.
- Spring Prompt overall: MiniMax M3.1 Flash 47, Mistral Large 4 33 (out of 100; how the score works).
- MiniMax M3.1 Flash is cheaper: $0.26 against $2.06 per million tokens, 7.9 times less.
- MiniMax M3.1 Flash scores higher on the Artificial Analysis Intelligence Index: 42.4 against 38.4.
- MiniMax M3.1 Flash writes faster: 159 tokens a second against 60.
- MiniMax M3.1 Flash is newer: released 19 Oct 2026, against 6 Oct 2026 for Mistral Large 4.
- On our own benchmarks, on the results both models have: MiniMax M3.1 Flash does better for product listings, decks from an analysis and marketing planning; Mistral Large 4 for user surveys.
- On third-party benchmarks, on the results both models have: MiniMax M3.1 Flash does better for reasoning and knowledge; Mistral Large 4 for agents and tool use.
Head to head
47
Spring Prompt overall
33
42.4
Intelligence Index
38.4
$0.26
Price per million tokens, blendedlower is better
$2.06
159
Output speed, tokens a second
60
1,048,576
Context window, tokens
524,288
Every fact side by side
| MiniMax M3.1 Flash | Mistral Large 4 | |
|---|---|---|
| Price per million tokens | $0.15 in · $0.60 out | $1.36 in · $4.18 out |
| Blended (3 in : 1 out) | $0.26 | $2.06 |
| Cheapest host, blended | – | – |
| Cached input, per million | $0.015 cached input | – |
| Intelligence Index | 42.4 | 38.4 |
| Spring Prompt overall | 47 | 33 |
| Output speed, tokens a second | 159 | 60 |
| Context window | 1,048,576 tokens | 524,288 tokens |
| Released | 19 Oct 2026 (newer) | 6 Oct 2026 |
| Weights | Closed | Closed |
| Developer based in | China | France |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.
Product listingsTurning a sparse product feed and photos into listings that can go live
7 of 24
21 of 24
Decks from an analysisTurning a finished analysis into a deck you could present as it is
14 of 24
23 of 24
User surveysPlanning a user survey and reading its results without being misled
13 of 13
7 of 13
Marketing planningPlanning a year of ad spend without overspending
12 of 24
14 of 24
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
32 of 168
28 of 168
Reasoning and knowledgeHard questions across science, maths and general knowledge
18 of 202
50 of 202
Speed under pressureGood decisions against a real clock (fast chess)
not measured
17 of 27
From our research
- MiniMax M3.1 Flash first look: careful with product data, unreliable on long jobs 19 Oct 2026
- AI Checkpoint: Sol’s new fast lane 9 Oct 2026
- Mistral Large 4 Is Fast, Cheap and Europe's Best Model. On Business Work, It Still Makes Things Up. 7 Oct 2026
- We Told 11 AI Models What Happened in 2026. Most of Them Thought We Were Making It Up. 12 Oct 2026
Every shared result
Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.
| Measure | MiniMax M3.1 Flash | Mistral Large 4 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 3 results | ||
| Cost per productUS dollars, lower is better 1 · 6 of 24 measured | $0 | $0.0033 |
| Reliably publish-ready% of products, higher is better 9 · 22 of 24 measured | 50.0% | 10.7%high reasoning |
| Reliably publish-ready · sales brief% of products, higher is better 6 · 19 of 24 measured | 37.5% | 1.8%high reasoning |
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better 1 · 6 of 24 measured | $0 | $0.0148 |
| Deck ratingrating, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 14 · 23 of 24 measured | 1,090 | 637 |
| Accurate decks% of tasks, higher is better 9 · 20 of 24 measured | 50.0% | 0.0% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better 12 · 14 of 24 measured | 40.3 | 37.0 |
| Audience scorescore out of 100, higher is better 12 · 14 of 24 measured | 60.8 | 55.8 |
| Business scorescore out of 100, higher is better 10 · 14 of 24 measured | 38.9 | 34.3 |
| SurveyBench measured by Spring Prompt · 3 results | ||
| Analysis ratingrating, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. 13 · 7 of 13 measured | 441 | 805 |
| Analysis scorepoints of 100, higher is better 13 · 8 of 13 measured | 7.90 | 63.8 |
| Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. 13 · 8 of 13 measured | 15.9% | 35.0% |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 21 · 36 of 249 measured | 42.4 | 38.4 |
| Humanity's Last Exam% of questions, higher is better 14 · 63 of 248 measured | 48.0% | 35.0% |
| Terminal-Bench 4.0% of tasks, higher is better 22 · 20 of 118 measured | 25.8% | 26.8% |
Show all 82 results (67 more)
| Measure | MiniMax M3.1 Flash | Mistral Large 4 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 28 results | ||
| Channel compliance% of products, higher is better 1 · 1 of 24 measured | 100.0% | 100.0% |
| Channel rules broken% of products, lower is better 12 · 23 of 24 measured | 0.6% | 3.0% |
| Channel rules broken · sales brief% of products, lower is better 1 · 18 of 24 measured | 0.0% | 1.2% |
| Claims to check% of products, lower is better 1 · 18 of 24 measured | 0.0% | 0.7%high reasoning |
| Claims to check · sales brief% of products, lower is better 1 · 17 of 24 measured | 0.0% | 3.3%high reasoning |
| Conflicts caught% of conflicts, higher is better 1 · 1 of 24 measured | 100.0% | 100.0% |
| Content quality% of checks, higher is better 17 · 14 of 24 measured | 95.8% | 96.5% |
| Content quality · sales brief% of checks, higher is better 17 · 18 of 24 measured | 96.0% | 95.9% |
| Decision accuracy% of decisions, higher is better 16 · 21 of 24 measured | 95.6% | 93.1% |
| Failed outputs% of products, lower is better 22 · 1 of 24 measured | 1.8% | 0.0% |
| Failed outputs · sales brief% of products, lower is better 22 · 1 of 24 measured | 1.8% | 0.0% |
| Field accuracy% of missing fields, higher is better 19 · 5 of 24 measured | 92.0% | 94.4% |
| Invented values% of filled values, lower is better 22 · 20 of 24 measured | 3.8% | 3.5% |
| Missing UK information% of products, lower is better 1 · 1 of 24 measured | 0.0% | 0.0% |
| Missing UK information · sales brief% of products, lower is better 1 · 1 of 24 measured | 0.0% | 0.0% |
| Not findable% of products, lower is better 16 · 9 of 24 measured | 7.1% | 4.2% |
| Not findable · sales brief% of products, lower is better 12 · 8 of 24 measured | 4.2% | 3.0%high reasoning |
| Publish-ready listings% of products, higher is better 11 · 20 of 24 measured | 63.1% | 39.9%high reasoning |
| Publish-ready listings · sales brief% of products, higher is better 7 · 19 of 24 measured | 57.1% | 10.7%high reasoning |
| Time per productseconds, lower is better 19 · 6 of 24 measured | 32.0 | 8.20 |
| Unsupported claimsclaims per product, lower is better 6 · 15 of 24 measured | 0.39 | 0.91high reasoning |
| Unsupported claims% of products, lower is better 5 · 17 of 24 measured | 7.7% | 33.3%high reasoning |
| Unsupported claims · sales briefclaims per product, lower is better 6 · 20 of 24 measured | 0.60 | 3.69high reasoning |
| Unsupported claims · sales brief% of products, lower is better 5 · 18 of 24 measured | 16.1% | 80.4%high reasoning |
| Wrong attributes% of products, lower is better 20 · 16 of 24 measured | 22.6% | 19.6%high reasoning |
| Wrong attributes · sales brief% of products, lower is better 20 · 11 of 24 measured | 21.4% | 17.3%high reasoning |
| Wrong category or variant% of products, lower is better 20 · 3 of 24 measured | 4.2% | 1.2% |
| Wrong category or variant · sales brief% of products, lower is better 21 · 5 of 24 measured | 4.8% | 1.8% |
| DeckBench measured by Spring Prompt · 14 results | ||
| Caveat dropped% of tasks, lower is better 24 · 12 of 24 measured | 33.3% | 16.7% |
| Clean layout% of tasks, higher is better 23 · 14 of 24 measured | 0.0% | 33.3% |
| Design quality% of the maximum, higher is better 21 · 24 of 24 measured | 46.2% | 30.5% |
| Draft figure quoted% of tasks, lower is better 12 · 9 of 24 measured | 33.3% | 16.7% |
| Findings missing% of tasks, lower is better 24 · 1 of 24 measured | 50.0% | 0.0% |
| Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 15 · 23 of 24 measured | 48.6% | 25.0% |
| Layout defects% of tasks, lower is better 23 · 14 of 24 measured | 100.0% | 66.7% |
| Misleading metric used% of tasks, lower is better 24 · 1 of 24 measured | 33.3% | 0.0% |
| Presentable decks% of tasks, higher is better 12 · 12 of 24 measured | 0.0% | 0.0% |
| Recommendation late or wrong% of tasks, lower is better 13 · 22 of 24 measured | 33.3% | 83.3% |
| Slides needing work% of tasks, lower is better 21 · 24 of 24 measured | 83.3% | 100.0% |
| Time per deckseconds, lower is better 20 · 15 of 24 measured | 330 | 114 |
| Unsupported claims% of tasks, lower is better 21 · 21 of 24 measured | 50.0% | 50.0% |
| Unsupported numbers% of tasks, lower is better 24 · 23 of 24 measured | 50.0% | 33.3% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better 17 · 18 of 24 measured | 17.3 | 16.7 |
| Contribution profitsimulated US dollars, higher is better 5 · 12 of 24 measured | $693,379 | $533,401 |
| Cost of a runUS dollars, lower is better 1 · 7 of 24 measured | $0 | $0.0657 |
| Months over budgetmonths of 12, lower is better 1 · 1 of 24 measured | 0 | 0high reasoning |
| Planning scorescore out of 100, higher is better 21 · 4 of 24 measured | 54.2 | 55.0high reasoning |
| Return on ad spendprofit per $1 spent, higher is better 9 · 11 of 24 measured | $0.52 | $0.45 |
| SurveyBench measured by Spring Prompt · 15 results | ||
| Leading or double-barrelled questions% of tasks, lower is better 12 · 4 of 13 measured | 77.8% | 33.3% |
| Mistaken next steps% of tasks, lower is better 13 · 11 of 13 measured | 100.0% | 66.7% |
| Numbers right% of answers, higher is better 1 · 8 of 13 measured | 100.0% | 98.9% |
| Objectives unanswerable% of tasks, lower is better 13 · 7 of 13 measured | 66.7% | 11.1% |
| Precision ignored% of tasks, lower is better 10 · 8 of 13 measured | 66.7% | 55.6% |
| Research scorepoints of 100, higher is better 13 · 8 of 13 measured | 16.4 | 71.2 |
| Sound research% of tasks, higher is better 6 · 6 of 13 measured | 0.0% | 0.0% |
| Stakeholder's view presumed% of tasks, lower is better 12 · 7 of 13 measured | 55.6% | 22.2% |
| Survey plan scorepoints of 100, higher is better 13 · 7 of 13 measured | 36.1 | 88.4 |
| Survey structure% of tasks, lower is better 13 · 3 of 13 measured | 77.8% | 11.1% |
| Traps handled% of traps, higher is better 8 · 7 of 13 measured | 75.0% | 78.7% |
| Traps missed% of tasks, lower is better 10 · 10 of 13 measured | 100.0% | 100.0% |
| Unsupported findings% of tasks, lower is better 10 · 10 of 13 measured | 100.0% | 100.0% |
| Wrong numbers% of tasks, lower is better 10 · 7 of 13 measured | 88.9% | 22.2% |
| Wrong recommendation% of tasks, lower is better 13 · 8 of 13 measured | 100.0% | 44.4% |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| Long-context reasoning (AA-LCR)% of questions, higher is better 45 · 30 of 239 measured | 79.7% | 81.3% |
| Output speedtokens per second, higher is better 21 · 61 of 78 measured | 159 | 60.2 |
| SciCode% of problems, higher is better 36 · 32 of 119 measured | 53.4% | 54.2% |
| Time to first answer tokenseconds, lower is better 51 · 68 of 78 measured | 13.3 | 33.9 |
Compare them with other models
- MiniMax vs Mistral AIEvery model from both labs, best against best
- MiniMax M3.1 Flash vs Claude Haiku 5.5Cheapest with similar intelligence
- MiniMax M3.1 Flash vs DeepSeek-V4.1-FlashFastest with similar intelligence
- MiniMax M3.1 Flash vs Claude Opus 5.5Most intelligent available today
- Mistral Large 4 vs Muse Spark 1.3Most intelligent at $1–3 per million tokens
- Mistral Large 4 vs DeepSeek-V4.1-FlashFastest with similar intelligence
- Mistral Large 4 vs Claude Opus 5.5Most intelligent available today
Quick answers
Is MiniMax M3.1 Flash better than Mistral Large 4?
Pick MiniMax M3.1 Flash: it scores higher (Spring Prompt overall 47 against 33) and costs less.
Which is cheaper, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash is cheaper: $0.26 against $2.06 per million tokens, 7.9 times less.
Which is faster, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash writes about 159 tokens a second on its usual API, against 60 for Mistral Large 4 (Artificial Analysis).
Which is better for product listings, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash ranks 7 of 24 models on sale for turning a sparse product feed and photos into listings that can go live, against 21 for Mistral Large 4.
Which is better for decks from an analysis, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash ranks 14 of 24 models on sale for turning a finished analysis into a deck you could present as it is, against 23 for Mistral Large 4.
Which is better for user surveys, MiniMax M3.1 Flash or Mistral Large 4?
Mistral Large 4 ranks 7 of 13 models on sale for planning a user survey and reading its results without being misled, against 13 for MiniMax M3.1 Flash.
Which is better for marketing planning, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash ranks 12 of 24 models on sale for planning a year of ad spend without overspending, against 14 for Mistral Large 4.
Which has the bigger context window, MiniMax M3.1 Flash or Mistral Large 4?
MiniMax M3.1 Flash: 1,048,576 tokens, against 524,288 for Mistral Large 4.
Run MiniMax M3.1 Flash and Mistral Large 4 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.