Compare / DeepSeek-V4-Pro (0423) vs MiniMax M3.1 Flash
DeepSeek-V4-Pro (0423) vs MiniMax M3.1 Flash
Which is cheaper, which is more capable, and which does better at each kind of work, from 31 results on 3 sources that measured both models.
Of the 31 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 MiniMax, every model from both labs.
DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash? The short answer
Pick MiniMax M3.1 Flash: it scores higher (Spring Prompt overall 47 against 31) and costs less.
- Spring Prompt overall: DeepSeek-V4-Pro (0423) 31, MiniMax M3.1 Flash 47 (out of 100; how the score works).
- MiniMax M3.1 Flash is cheaper: $0.26 against $1.77 per million tokens, 6.7 times less.
- MiniMax M3.1 Flash scores higher on the Artificial Analysis Intelligence Index: 42.4 against 30.4.
- MiniMax M3.1 Flash is newer: released 19 Oct 2026, against 24 Apr 2026 for DeepSeek-V4-Pro (0423).
- On our own benchmarks, on the results both models have: MiniMax M3.1 Flash does better for decks from an analysis and marketing planning.
- On third-party benchmarks, on the results both models have: MiniMax M3.1 Flash does better for agents and tool use and reasoning and knowledge.
Head to head
Every fact side by side
| DeepSeek-V4-Pro (0423) | MiniMax M3.1 Flash | |
|---|---|---|
| Price per million tokens | $1.42 in · $2.83 out | $0.15 in · $0.60 out |
| Blended (3 in : 1 out) | $1.77 | $0.26 |
| Cheapest host, blended | $1.19 (StreamLake, FP8) | – |
| Cached input, per million | – | $0.015 cached input |
| Intelligence Index | 30.4 | 42.4 |
| Spring Prompt overall | 31 | 47 |
| Output speed, tokens a second | – | 159 |
| Context window | 1,048,576 tokens | 1,048,576 tokens |
| Released | 24 Apr 2026 | 19 Oct 2026 (newer) |
| Weights | Open | Closed |
| Developer based in | China | China |
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.
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 | DeepSeek-V4-Pro (0423) | MiniMax M3.1 Flash |
|---|---|---|
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better 11 · 1 of 24 measured | $0.0465 | $0 |
| 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. 24 · 14 of 24 measured | 599 | 1,090 |
| Accurate decks% of tasks, higher is better 20 · 9 of 24 measured | 0.0% | 50.0% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better 17 · 12 of 24 measured | 28.3 | 40.3 |
| Audience scorescore out of 100, higher is better 20 · 12 of 24 measured | 44.8 | 60.8 |
| Business scorescore out of 100, higher is better 17 · 10 of 24 measured | 23.8 | 38.9 |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 53 · 21 of 249 measured | 30.4max reasoning | 42.4 |
| Humanity's Last Exam% of questions, higher is better 55 · 14 of 248 measured | 37.5%max reasoning | 48.0% |
| Terminal-Bench 4.0% of tasks, higher is better 30 · 22 of 118 measured | 14.6%max reasoning | 25.8% |
Show all 31 results (22 more)
| Measure | DeepSeek-V4-Pro (0423) | MiniMax M3.1 Flash |
|---|---|---|
| DeckBench measured by Spring Prompt · 14 results | ||
| Caveat dropped% of tasks, lower is better 12 · 24 of 24 measured | 16.7% | 33.3% |
| Clean layout% of tasks, higher is better 14 · 23 of 24 measured | 33.3% | 0.0% |
| Design quality% of the maximum, higher is better 23 · 21 of 24 measured | 40.9% | 46.2% |
| Draft figure quoted% of tasks, lower is better 12 · 12 of 24 measured | 33.3% | 33.3% |
| Findings missing% of tasks, lower is better 17 · 24 of 24 measured | 16.7% | 50.0% |
| 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. 24 · 15 of 24 measured | 24.4% | 48.6% |
| Layout defects% of tasks, lower is better 14 · 23 of 24 measured | 66.7% | 100.0% |
| Misleading metric used% of tasks, lower is better 1 · 24 of 24 measured | 0.0% | 33.3% |
| 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 20 · 13 of 24 measured | 66.7% | 33.3% |
| Slides needing work% of tasks, lower is better 21 · 21 of 24 measured | 83.3% | 83.3% |
| Time per deckseconds, lower is better 21 · 20 of 24 measured | 332 | 330 |
| Unsupported claims% of tasks, lower is better 24 · 21 of 24 measured | 66.7% | 50.0% |
| Unsupported numbers% of tasks, lower is better 1 · 24 of 24 measured | 0.0% | 50.0% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better 22 · 17 of 24 measured | 9.1 | 17.3 |
| Contribution profitsimulated US dollars, higher is better 16 · 5 of 24 measured | $301,289 | $693,379 |
| Cost of a runUS dollars, lower is better 13 · 1 of 24 measured | $0.20 | $0 |
| Months over budgetmonths of 12, lower is better 1 · 1 of 24 measured | 0 | 0 |
| Planning scorescore out of 100, higher is better 22 · 21 of 24 measured | 54.0 | 54.2 |
| Return on ad spendprofit per $1 spent, higher is better 16 · 9 of 24 measured | $0.26 | $0.52 |
| Artificial Analysis reported by Artificial Analysis · 2 results | ||
| Long-context reasoning (AA-LCR)% of questions, higher is better 84 · 45 of 239 measured | 74.7%max reasoning | 79.7% |
| SciCode% of problems, higher is better 48 · 36 of 119 measured | 50.8%max reasoning | 53.4% |
Compare them with other models
- DeepSeek vs MiniMaxEvery model from both labs, best against best
- DeepSeek-V4-Pro (0423) vs Muse Spark 1.3Most intelligent at $1–3 per million tokens
- DeepSeek-V4-Pro (0423) vs Claude Opus 5.5Most intelligent available today
- 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
Quick answers
Is DeepSeek-V4-Pro (0423) better than MiniMax M3.1 Flash?
Pick MiniMax M3.1 Flash: it scores higher (Spring Prompt overall 47 against 31) and costs less.
Which is cheaper, DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash?
MiniMax M3.1 Flash is cheaper: $0.26 against $1.77 per million tokens, 6.7 times less.
Which is better for decks from an analysis, DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash?
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 24 for DeepSeek-V4-Pro (0423).
Which is better for marketing planning, DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash?
MiniMax M3.1 Flash ranks 12 of 24 models on sale for planning a year of ad spend without overspending, against 17 for DeepSeek-V4-Pro (0423).
Which is better for agents and tool use, DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash?
MiniMax M3.1 Flash ranks 32 of 168 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 91 for DeepSeek-V4-Pro (0423).
Which is better for reasoning and knowledge, DeepSeek-V4-Pro (0423) or MiniMax M3.1 Flash?
MiniMax M3.1 Flash ranks 18 of 202 models on sale for hard questions across science, maths and general knowledge, against 48 for DeepSeek-V4-Pro (0423).
Run DeepSeek-V4-Pro (0423) and MiniMax M3.1 Flash on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.