Compare / Mistral Large 4 vs Particle-1
Mistral Large 4 vs Particle-1
Which is cheaper, which is more capable, and which does better at each kind of work, from 75 results on 4 sources that measured both models.
Mistral AIMistral Large 4
Particle AIParticle-1
Mistral Large 4 wins 1413 too close to callParticle-1 wins 48
Of the 75 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 5 November 2026 (catalogue release 2026-11-05-44c5fb63d296).
Mistral Large 4 or Particle-1? The short answer
They score about the same, so pick Particle-1: it costs 2.8 times less per product listing on CatalogBench.
- Spring Prompt overall: Mistral Large 4 32, Particle-1 33 (out of 100; how the score works).
- Particle-1 is cheaper to run: $0.001 against $0.003 per product listing on CatalogBench (2.8 times less).
- On our own benchmarks, on the results both models have: Mistral Large 4 does better for product listings; Particle-1 for decks from an analysis, user surveys and marketing planning.
Head to head
32
Spring Prompt overall
33
$2.06
Price per million tokens, blendedlower is better
$0.16
1,048,576
Context window, tokens
1,048,576
Every fact side by side
| Mistral Large 4 | Particle-1 | |
|---|---|---|
| Price per million tokens | $1.36 in · $4.18 out | $0.050 in · $0.50 out |
| Blended (3 in : 1 out) | $2.06 | $0.16 |
| Cheapest host, blended | $1.05 (Mistral) | – |
| Cost per product listing (CatalogBench) | $0.003 | $0.001 |
| Intelligence Index | 38.4 | – |
| Spring Prompt overall | 32 | 33 |
| Output speed, tokens a second | 60 | – |
| Context window | 1,048,576 tokens | 1,048,576 tokens |
| Released | 6 Oct 2026 | – |
| Weights | Open | Closed |
| Developer based in | 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
22 of 27
26 of 27
Decks from an analysisTurning a finished analysis into a deck you could present as it is
26 of 28
15 of 28
User surveysPlanning a user survey and reading its results without being misled
10 of 17
7 of 17
Marketing planningPlanning a year of ad spend without overspending
18 of 28
14 of 28
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
28 of 160
not measured
Reasoning and knowledgeHard questions across science, maths and general knowledge
52 of 188
not measured
Speed under pressureGood decisions against a real clock (fast chess)
22 of 31
32 of 32
Their slides, side by side
The same brief for both: Manchester store investment review for Hearthside Coffee, an invented company. First slide of each deck from DeckBench.
From our research
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 | Mistral Large 4 | Particle-1 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 3 results | ||
| Cost per productUS dollars, lower is better 7 · 3 of 28 measured | $0.0033 | $0.0012 |
| Reliably publish-ready% of products, higher is better 23 · 26 of 27 measured | 10.7%high reasoning | 0.0% |
| Reliably publish-ready · sales brief% of products, higher is better 20 · 22 of 27 measured | 1.8%high reasoning | 0.0% |
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better 7 · 2 of 28 measured | $0.0148 | $0.0086 |
| 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. 26 · 15 of 28 measured | 643 | 1,163 |
| Accurate decks% of tasks, higher is better 23 · 19 of 28 measured | 0.0% | 16.7% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better 18 · 14 of 28 measured | 37.0 | 43.3 |
| Audience scorescore out of 100, higher is better 18 · 13 of 28 measured | 55.8 | 65.1 |
| Business scorescore out of 100, higher is better 18 · 13 of 28 measured | 34.3 | 39.0 |
| 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. 10 · 7 of 17 measured | 831 | 1,193 |
| Cost per taskUS dollars, lower is better 11 · 1 of 17 measured | $0.0924 | $0.0140 |
| Analysis scorepoints of 100, higher is better 10 · 6 of 17 measured | 63.8 | 88.4 |
Show all 75 results (63 more)
| Measure | Mistral Large 4 | Particle-1 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 26 results | ||
| Channel compliance% of products, higher is better 1 · 1 of 28 measured | 100.0% | 100.0% |
| Channel rules broken% of products, lower is better 27 · 15 of 28 measured | 3.0% | 0.6% |
| Channel rules broken · sales brief% of products, lower is better 21 · 14 of 28 measured | 1.2% | 0.6% |
| Conflicts caught% of conflicts, higher is better 1 · 1 of 28 measured | 100.0% | 100.0% |
| Content quality% of checks, higher is better 17 · 13 of 28 measured | 96.5% | 97.1% |
| Content quality · sales brief% of checks, higher is better 20 · 19 of 28 measured | 95.9% | 96.1% |
| Decision accuracy% of decisions, higher is better 25 · 13 of 28 measured | 93.1% | 96.8% |
| Failed outputs% of products, lower is better 1 · 1 of 28 measured | 0.0% | 0.0% |
| Failed outputs · sales brief% of products, lower is better 1 · 1 of 28 measured | 0.0% | 0.0% |
| Field accuracy% of missing fields, higher is better 6 · 13 of 28 measured | 94.4% | 93.7% |
| Invented values% of filled values, lower is better 24 · 14 of 28 measured | 3.5% | 1.7% |
| Missing UK information% of products, lower is better 1 · 1 of 28 measured | 0.0% | 0.0% |
| Missing UK information · sales brief% of products, lower is better 1 · 1 of 28 measured | 0.0% | 0.0% |
| Not findable% of products, lower is better 12 · 1 of 28 measured | 4.2% | 0.0% |
| Not findable · sales brief% of products, lower is better 11 · 1 of 28 measured | 3.0%high reasoning | 0.0% |
| Publish-ready listings% of products, higher is better 22 · 27 of 28 measured | 39.9%high reasoning | 0.0% |
| Publish-ready listings · sales brief% of products, higher is better 21 · 26 of 28 measured | 10.7%high reasoning | 0.0% |
| Time per productseconds, lower is better 6 · 20 of 28 measured | 8.20 | 23.5 |
| Unsupported claims% of products, lower is better 21 · 15 of 28 measured | 33.3%high reasoning | 19.6% |
| Unsupported claimsclaims per product, lower is better 18 · 16 of 28 measured | 0.91high reasoning | 0.71 |
| Unsupported claims · sales briefclaims per product, lower is better 23 · 17 of 28 measured | 3.69high reasoning | 2.36 |
| Unsupported claims · sales brief% of products, lower is better 22 · 12 of 28 measured | 80.4%high reasoning | 49.4% |
| Wrong attributes% of products, lower is better 19 · 19 of 28 measured | 19.6%high reasoning | 19.6% |
| Wrong attributes · sales brief% of products, lower is better 13 · 14 of 28 measured | 17.3%high reasoning | 17.9% |
| Wrong category or variant% of products, lower is better 3 · 5 of 28 measured | 1.2% | 1.8% |
| Wrong category or variant · sales brief% of products, lower is better 5 · 5 of 28 measured | 1.8% | 1.8% |
| DeckBench measured by Spring Prompt · 14 results | ||
| Caveat dropped% of tasks, lower is better 14 · 1 of 28 measured | 16.7% | 0.0% |
| Clean layout% of tasks, higher is better 15 · 22 of 28 measured | 33.3% | 16.7% |
| Design quality% of the maximum, higher is better 27 · 22 of 28 measured | 30.5% | 54.4% |
| Draft figure quoted% of tasks, lower is better 12 · 21 of 28 measured | 16.7% | 50.0% |
| Findings missing% of tasks, lower is better 1 · 1 of 28 measured | 0.0% | 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. 26 · 15 of 28 measured | 26.0% | 51.7% |
| Layout defects% of tasks, lower is better 15 · 22 of 28 measured | 66.7% | 83.3% |
| Misleading metric used% of tasks, lower is better 1 · 1 of 28 measured | 0.0% | 0.0% |
| Presentable decks% of tasks, higher is better 13 · 13 of 28 measured | 0.0% | 0.0% |
| Recommendation late or wrong% of tasks, lower is better 25 · 16 of 28 measured | 83.3% | 33.3% |
| Slides needing work% of tasks, lower is better 28 · 8 of 28 measured | 100.0% | 33.3% |
| Time per deckseconds, lower is better 15 · 17 of 28 measured | 114 | 154 |
| Unsupported claims% of tasks, lower is better 25 · 1 of 28 measured | 50.0% | 0.0% |
| Unsupported numbers% of tasks, lower is better 27 · 1 of 28 measured | 33.3% | 0.0% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better 22 · 15 of 28 measured | 16.7 | 27.8 |
| Contribution profitsimulated US dollars, higher is better 16 · 9 of 28 measured | $533,401 | $684,624 |
| Cost of a runUS dollars, lower is better 8 · 3 of 28 measured | $0.0657 | $0.0234 |
| Months over budgetmonths of 12, lower is better 1 · 1 of 28 measured | 0high reasoning | 0 |
| Planning scorescore out of 100, higher is better 5 · 25 of 28 measured | 55.0high reasoning | 54.3 |
| Return on ad spendprofit per $1 spent, higher is better 15 · 13 of 28 measured | $0.45 | $0.51 |
| SurveyBench measured by Spring Prompt · 17 results | ||
| 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. 12 · 8 of 17 measured | 35.1% | 50.0% |
| Leading or double-barrelled questions% of tasks, lower is better 7 · 3 of 17 measured | 33.3% | 11.1% |
| Mistaken next steps% of tasks, lower is better 14 · 6 of 17 measured | 66.7% | 22.2% |
| Numbers right% of answers, higher is better 11 · 1 of 17 measured | 98.9% | 100.0% |
| Objectives unanswerable% of tasks, lower is better 8 · 8 of 17 measured | 11.1% | 11.1% |
| Precision ignored% of tasks, lower is better 13 · 1 of 17 measured | 55.6% | 0.0% |
| Research scorepoints of 100, higher is better 10 · 7 of 17 measured | 71.2 | 89.7 |
| Sound research% of tasks, higher is better 9 · 6 of 17 measured | 0.0% | 22.2% |
| Stakeholder's view presumed% of tasks, lower is better 10 · 1 of 17 measured | 22.2% | 0.0% |
| Survey plan scorepoints of 100, higher is better 10 · 6 of 17 measured | 88.4 | 92.6 |
| Survey structure% of tasks, lower is better 3 · 12 of 17 measured | 11.1% | 33.3% |
| Time per taskseconds, lower is better 16 · 13 of 17 measured | 439 | 252 |
| Traps handled% of traps, higher is better 11 · 2 of 17 measured | 78.7% | 95.7% |
| Traps missed% of tasks, lower is better 15 · 3 of 17 measured | 100.0% | 33.3% |
| Unsupported findings% of tasks, lower is better 15 · 7 of 17 measured | 100.0% | 55.6% |
| Wrong numbers% of tasks, lower is better 9 · 1 of 17 measured | 22.2% | 0.0% |
| Wrong recommendation% of tasks, lower is better 10 · 3 of 17 measured | 44.4% | 11.1% |
Compare them with other models
- 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
- Particle-1 vs Claude Haiku 5.5Highest-scoring overall at under $0.25 per million tokens
- Particle-1 vs Claude Opus 5.5Highest-scoring overall available today
- Particle-1 vs GLM-5.3 FlashX LegalJust above overall
Quick answers
Is Mistral Large 4 better than Particle-1?
They score about the same, so pick Particle-1: it costs 2.8 times less per product listing on CatalogBench.
Which is cheaper, Mistral Large 4 or Particle-1?
Particle-1 is cheaper to run: $0.001 against $0.003 per product listing on CatalogBench (2.8 times less).
Which is better for product listings, Mistral Large 4 or Particle-1?
Mistral Large 4 ranks 22 of 27 models on sale for turning a sparse product feed and photos into listings that can go live, against 26 for Particle-1.
Which is better for decks from an analysis, Mistral Large 4 or Particle-1?
Particle-1 ranks 15 of 28 models on sale for turning a finished analysis into a deck you could present as it is, against 26 for Mistral Large 4.
Which is better for user surveys, Mistral Large 4 or Particle-1?
Particle-1 ranks 7 of 17 models on sale for planning a user survey and reading its results without being misled, against 10 for Mistral Large 4.
Which is better for marketing planning, Mistral Large 4 or Particle-1?
Particle-1 ranks 14 of 28 models on sale for planning a year of ad spend without overspending, against 18 for Mistral Large 4.
Which is better for speed under pressure, Mistral Large 4 or Particle-1?
Mistral Large 4 ranks 22 of 31 models on sale for good decisions against a real clock (fast chess), against 32 for Particle-1.
Run Mistral Large 4 and Particle-1 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.