Compare / DeepSeek-V4.1-Flash-1017 vs Mistral Medium 3.5
DeepSeek-V4.1-Flash-1017 vs Mistral Medium 3.5
Which is cheaper, which is more capable, and which does better at each kind of work, from 89 results on 6 sources that measured both models.
DeepSeekDeepSeek-V4.1-Flash-1017
Mistral AIMistral Medium 3.5
DeepSeek-V4.1-Flash-1017 wins 5912 too close to callMistral Medium 3.5 wins 18
Of the 89 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 DeepSeek vs Mistral AI, every model from both labs.
DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5? The short answer
Pick DeepSeek-V4.1-Flash-1017: it scores higher (Spring Prompt overall 44 against 16) and costs less.
- Spring Prompt overall: DeepSeek-V4.1-Flash-1017 44, Mistral Medium 3.5 16 (out of 100; how the score works).
- DeepSeek-V4.1-Flash-1017 is cheaper to run: $0.002 against $0.008 per product listing on CatalogBench (5.2 times less).
- DeepSeek-V4.1-Flash-1017 scores higher on the Artificial Analysis Intelligence Index: 45.1 against 14.2.
- DeepSeek-V4.1-Flash-1017 writes faster: 212 tokens a second against 157.
- DeepSeek-V4.1-Flash-1017 is newer: released 17 Oct 2026, against 28 Apr 2026 for Mistral Medium 3.5.
- On our own benchmarks, on the results both models have: DeepSeek-V4.1-Flash-1017 does better for decks from an analysis, user surveys and marketing planning.
- On third-party benchmarks, on the results both models have: DeepSeek-V4.1-Flash-1017 does better for agents and tool use and reasoning and knowledge.
Head to head
44
Spring Prompt overall
16
45.1
Intelligence Index
14.2
$0.52
Price per million tokens, blendedlower is better
$3.00
212
Output speed, tokens a second
157
1,048,576
Context window, tokens
262,144
Every fact side by side
| DeepSeek-V4.1-Flash-1017 | Mistral Medium 3.5 | |
|---|---|---|
| Price per million tokens | $0.30 in · $1.20 out | $1.50 in · $7.50 out |
| Blended (3 in : 1 out) | $0.52 | $3.00 |
| Cheapest host, blended | – | $3.00 (Mistral) |
| Cost per product listing (CatalogBench) | $0.002 | $0.008 |
| Cached input, per million | $0.006 cached input | $0.15 cached input |
| Intelligence Index | 45.1 | 14.2 |
| Spring Prompt overall | 44 | 16 |
| Output speed, tokens a second | 212 | 157 |
| Context window | 1,048,576 tokens | 262,144 tokens |
| Released | 17 Oct 2026 (newer) | 28 Apr 2026 |
| Weights | Open | 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
21 of 24
18 of 24
Decks from an analysisTurning a finished analysis into a deck you could present as it is
12 of 24
22 of 24
User surveysPlanning a user survey and reading its results without being misled
5 of 13
11 of 13
Marketing planningPlanning a year of ad spend without overspending
11 of 24
24 of 24
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
22 of 168
97 of 168
Reasoning and knowledgeHard questions across science, maths and general knowledge
22 of 202
126 of 202
WritingWhat people prefer in blind comparisons, and judged writing quality
not measured
87 of 189
Sticking to the factsSummarising without inventing things, and factual answers
not measured
101 of 153
Speed under pressureGood decisions against a real clock (fast chess)
21 of 27
9 of 27
From our research
- DeepSeek-V4.1-Flash-1017 review: better at analysis, much worse at sticking to the product data 19 Oct 2026
- Mercury 3 first look: very fast with thinking off, too slow with it on 20 Oct 2026
- MiniMax M3.1 Flash first look: careful with product data, unreliable on long jobs 19 Oct 2026
- Mistral Large 4 Is Fast, Cheap and Europe's Best Model. On Business Work, It Still Makes Things Up. 7 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 | DeepSeek-V4.1-Flash-1017 | Mistral Medium 3.5 |
|---|---|---|
| BulletBench measured by Spring Prompt · 3 results | ||
| Cost per game · Bullet 60sUS dollars, lower is better 13 · 28 of 32 measured | $0.0047low reasoning | $0.0354no reasoning |
| Ladder Elo · Bullet 60sladder Elo, higher is better 27 · 8 of 32 measured | 147low reasoning | 572no reasoning |
| Games lost on time · Bullet 60s% of games, lower is better 29 · 1 of 32 measured | 87.5%low reasoning | 0.0%no reasoning |
| CatalogBench measured by Spring Prompt · 3 results | ||
| Cost per productUS dollars, lower is better 4 · 10 of 24 measured | $0.0015 | $0.0080 |
| Reliably publish-ready% of products, higher is better 19 · 20 of 24 measured | 21.4% | 17.9% |
| Reliably publish-ready · sales brief% of products, higher is better 21 · 12 of 24 measured | 0.0% | 10.7% |
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better 7 · 8 of 24 measured | $0.0255 | $0.0448 |
| 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. 12 · 22 of 24 measured | 1,145 | 656 |
| Accurate decks% of tasks, higher is better 14 · 20 of 24 measured | 33.3% | 0.0% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better 11 · 24 of 24 measured | 41.4 | 15.2 |
| Audience scorescore out of 100, higher is better 9 · 24 of 24 measured | 65.3 | 28.5 |
| Business scorescore out of 100, higher is better 11 · 24 of 24 measured | 38.3 | 3.4 |
| 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. 5 · 11 of 13 measured | 1,221 | 597 |
| Cost per taskUS dollars, lower is better 4 · 6 of 12 measured | $0.0373 | $0.0730 |
| Analysis scorepoints of 100, higher is better 4 · 10 of 13 measured | 90.2 | 56.7 |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 16 · 139 of 249 measured | 45.1max reasoning | 14.2 |
| Humanity's Last Exam% of questions, higher is better 27 · 137 of 248 measured | 43.3%max reasoning | 13.8% |
| Terminal-Bench 4.0% of tasks, higher is better 16 · 70 of 118 measured | 33.8%max reasoning | 0.0% |
Show all 89 results (71 more)
| Measure | DeepSeek-V4.1-Flash-1017 | Mistral Medium 3.5 |
|---|---|---|
| BulletBench measured by Spring Prompt · 2 results | ||
| Invalid moves · Bullet 60s% of moves, lower is better 1 · 28 of 32 measured | 0.0%low reasoning | 0.4%no reasoning |
| Median move time · Bullet 60smilliseconds, lower is better 31 · 6 of 32 measured | 3.3 slow reasoning | 0.5 sno reasoning |
| 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 19 · 1 of 24 measured | 1.2% | 0.0% |
| Channel rules broken · sales brief% of products, lower is better 22 · 18 of 24 measured | 2.4% | 1.2% |
| Claims to check% of products, lower is better 19 · 22 of 24 measured | 1.2% | 3.6% |
| Claims to check · sales brief% of products, lower is better 18 · 22 of 24 measured | 3.6% | 25.6% |
| Conflicts caught% of conflicts, higher is better 1 · 23 of 24 measured | 100.0% | 90.9% |
| Content quality% of checks, higher is better 20 · 23 of 24 measured | 95.3% | 94.1% |
| Content quality · sales brief% of checks, higher is better 19 · 21 of 24 measured | 94.8% | 94.1% |
| Decision accuracy% of decisions, higher is better 18 · 22 of 24 measured | 94.9% | 92.6% |
| Failed outputs% of products, lower is better 1 · 1 of 24 measured | 0.0% | 0.0% |
| Failed outputs · sales brief% of products, lower is better 1 · 1 of 24 measured | 0.0% | 0.0% |
| Field accuracy% of missing fields, higher is better 10 · 20 of 24 measured | 93.7% | 91.7% |
| Invented values% of filled values, lower is better 13 · 1 of 24 measured | 1.7% | 0.0% |
| 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 5 · 23 of 24 measured | 3.6% | 14.3% |
| Not findable · sales brief% of products, lower is better 7 · 21 of 24 measured | 2.4% | 10.1% |
| Publish-ready listings% of products, higher is better 19 · 22 of 24 measured | 40.5% | 29.8% |
| Publish-ready listings · sales brief% of products, higher is better 21 · 15 of 24 measured | 8.3% | 17.3% |
| Time per productseconds, lower is better 3 · 2 of 24 measured | 6.70 | 4.90 |
| Unsupported claimsclaims per product, lower is better 21 · 22 of 24 measured | 1.70 | 1.74 |
| Unsupported claims% of products, lower is better 20 · 22 of 24 measured | 44.6% | 49.4% |
| Unsupported claims · sales briefclaims per product, lower is better 22 · 18 of 24 measured | 4.01 | 3.55 |
| Unsupported claims · sales brief% of products, lower is better 23 · 16 of 24 measured | 85.7% | 76.8% |
| Wrong attributes% of products, lower is better 13 · 22 of 24 measured | 17.9% | 31.0% |
| Wrong attributes · sales brief% of products, lower is better 21 · 23 of 24 measured | 22.0% | 36.3% |
| Wrong category or variant% of products, lower is better 2 · 24 of 24 measured | 0.6% | 11.3% |
| Wrong category or variant · sales brief% of products, lower is better 3 · 24 of 24 measured | 1.2% | 13.1% |
| DeckBench measured by Spring Prompt · 14 results | ||
| Caveat dropped% of tasks, lower is better 12 · 1 of 24 measured | 16.7% | 0.0% |
| Clean layout% of tasks, higher is better 20 · 6 of 24 measured | 16.7% | 66.7% |
| Design quality% of the maximum, higher is better 9 · 22 of 24 measured | 68.5% | 43.7% |
| Draft figure quoted% of tasks, lower is better 19 · 19 of 24 measured | 50.0% | 50.0% |
| Findings missing% of tasks, lower is better 1 · 1 of 24 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. 14 · 22 of 24 measured | 53.3% | 25.8% |
| Layout defects% of tasks, lower is better 20 · 6 of 24 measured | 83.3% | 33.3% |
| Misleading metric used% of tasks, lower is better 1 · 1 of 24 measured | 0.0% | 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 9 · 17 of 24 measured | 16.7% | 50.0% |
| Slides needing work% of tasks, lower is better 16 · 11 of 24 measured | 66.7% | 50.0% |
| Time per deckseconds, lower is better 7 · 2 of 24 measured | 73.0 | 25.0 |
| Unsupported claims% of tasks, lower is better 11 · 17 of 24 measured | 16.7% | 33.3% |
| Unsupported numbers% of tasks, lower is better 1 · 1 of 24 measured | 0.0% | 0.0% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better 15 · 23 of 24 measured | 20.2 | 8.7 |
| Contribution profitsimulated US dollars, higher is better 10 · 23 of 24 measured | $585,572 | −$159,882 |
| Cost of a runUS dollars, lower is better 8 · 10 of 24 measured | $0.0799 | $0.15 |
| Months over budgetmonths of 12, lower is better 1 · 24 of 24 measured | 0 | 3 |
| Planning scorescore out of 100, higher is better 20 · 2 of 24 measured | 54.3 | 55.2 |
| Return on ad spendprofit per $1 spent, higher is better 12 · 23 of 24 measured | $0.44 | −$0.17 |
| 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. 5 · 11 of 13 measured | 60.2% | 24.3% |
| Leading or double-barrelled questions% of tasks, lower is better 3 · 8 of 13 measured | 11.1% | 55.6% |
| Mistaken next steps% of tasks, lower is better 1 · 12 of 13 measured | 0.0% | 77.8% |
| Numbers right% of answers, higher is better 1 · 10 of 13 measured | 100.0% | 75.5% |
| Objectives unanswerable% of tasks, lower is better 7 · 7 of 13 measured | 11.1% | 11.1% |
| Precision ignored% of tasks, lower is better 1 · 11 of 13 measured | 0.0% | 100.0% |
| Research scorepoints of 100, higher is better 4 · 10 of 13 measured | 90.9 | 64.2 |
| Sound research% of tasks, higher is better 3 · 6 of 13 measured | 33.3% | 0.0% |
| Stakeholder's view presumed% of tasks, lower is better 1 · 11 of 13 measured | 0.0% | 33.3% |
| Survey plan scorepoints of 100, higher is better 4 · 11 of 13 measured | 92.5 | 81.5 |
| Survey structure% of tasks, lower is better 10 · 3 of 13 measured | 33.3% | 11.1% |
| Time per taskseconds, lower is better 6 · 2 of 12 measured | 134 | 47.5 |
| Traps handled% of traps, higher is better 5 · 12 of 13 measured | 91.3% | 59.4% |
| Traps missed% of tasks, lower is better 5 · 10 of 13 measured | 44.4% | 100.0% |
| Unsupported findings% of tasks, lower is better 4 · 10 of 13 measured | 44.4% | 100.0% |
| Wrong numbers% of tasks, lower is better 1 · 11 of 13 measured | 0.0% | 100.0% |
| Wrong recommendation% of tasks, lower is better 3 · 8 of 13 measured | 11.1% | 44.4% |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| Long-context reasoning (AA-LCR)% of questions, higher is better 19 · 118 of 239 measured | 82.7%max reasoning | 69.3% |
| Output speedtokens per second, higher is better 15 · 22 of 78 measured | 212max reasoning | 157 |
| SciCode% of problems, higher is better 24 · 83 of 119 measured | 55.9%max reasoning | 40.2% |
| Time to first answer tokenseconds, lower is better 46 · 52 of 78 measured | 10.5max reasoning | 13.3 |
Compare them with other models
- DeepSeek vs Mistral AIEvery model from both labs, best against best
- DeepSeek-V4.1-Flash-1017 vs Claude Haiku 5.5Cheapest with similar intelligence
- DeepSeek-V4.1-Flash-1017 vs Muse Spark 1.3Fastest with similar intelligence
- DeepSeek-V4.1-Flash-1017 vs Claude Opus 5.5Most intelligent available today
- Mistral Medium 3.5 vs Claude Opus 5.5Most intelligent at $3–10 per million tokens
- Mistral Medium 3.5 vs Mistral Large 4Most intelligent from the EU
Quick answers
Is DeepSeek-V4.1-Flash-1017 better than Mistral Medium 3.5?
Pick DeepSeek-V4.1-Flash-1017: it scores higher (Spring Prompt overall 44 against 16) and costs less.
Which is cheaper, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017 is cheaper to run: $0.002 against $0.008 per product listing on CatalogBench (5.2 times less).
Which is faster, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017 writes about 212 tokens a second on its usual API, against 157 for Mistral Medium 3.5 (Artificial Analysis).
Which is better for product listings, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
Mistral Medium 3.5 ranks 18 of 24 models on sale for turning a sparse product feed and photos into listings that can go live, against 21 for DeepSeek-V4.1-Flash-1017.
Which is better for decks from an analysis, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017 ranks 12 of 24 models on sale for turning a finished analysis into a deck you could present as it is, against 22 for Mistral Medium 3.5.
Which is better for user surveys, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017 ranks 5 of 13 models on sale for planning a user survey and reading its results without being misled, against 11 for Mistral Medium 3.5.
Which is better for marketing planning, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017 ranks 11 of 24 models on sale for planning a year of ad spend without overspending, against 24 for Mistral Medium 3.5.
Which has the bigger context window, DeepSeek-V4.1-Flash-1017 or Mistral Medium 3.5?
DeepSeek-V4.1-Flash-1017: 1,048,576 tokens, against 262,144 for Mistral Medium 3.5.
Run DeepSeek-V4.1-Flash-1017 and Mistral Medium 3.5 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.