MiniMax M3 vs DeepSeek V4 Flash
Market Position
LiveBench Quality Versus Estimated Output Cost
This current-market view appears only when both compared models are reviewed current models with publisher-reported LiveBench token accounting and current sourced USD output rates.
Model Markets Rankings
Intelligence, Cost, and Efficiency
| Ranking | MiniMax-M3 | DeepSeek-V4-Flash |
|---|---|---|
| IntelligenceHigher is better · MM Intelligence v1.2 | UnrankedNot in the 22-model eligible cohort | #19 of 2241.8 score |
| CostLower is better · Published-token output estimate | #4 of 36$0.017 per LiveBench case | #2 of 36$0.0060 per LiveBench case |
| EfficiencyHigher is better · MM Efficiency v1.0 | UnrankedNot in the 22-model eligible cohort | #1 of 2270.9 score |
Ranks come from the current complete eligible cohorts. Green highlights appear only when both models are ranked in the same metric. Missing required inputs remain unranked, and the three dimensions are not collapsed into an overall winner.
Benchmark Performance
Available Benchmarks
| Benchmark | MiniMax-M3 | DeepSeek-V4-Flash |
|---|---|---|
| LMArena Text Arenatext-2026-09-01-011508720696 · arena_rating · statistical tie | 1,435.20100% of row best · rating · minimax-m3; 95% CI [1430.86827835, 1439.53977389]; votes 44544; rank 77 | 1,431.98100% of row best · rating · deepseek-v4-flash; 95% CI [1427.84091263, 1436.10937297]; votes 48963; rank 80 |
| LiveBench2026-06-25 · overall · leader | 70.26100% of row best · percent · minimax-m3 · 15,530 output tokens / case | 69.6799% of row best · percent · deepseek-v4-flash · 34,434 output tokens / case |
| Overall ResultCounted from the protocol-matched rows above · 1 tie | 1 benchmark winOverall lead | 0 benchmark wins |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Technical Differences
Side-by-Side Facts
| Field | MiniMax-M3 | DeepSeek-V4-Flash |
|---|---|---|
| Developer | MiniMax | DeepSeek |
| Family | Minimax M3 | Deepseek V4 Flash |
| Model | MiniMax-M3 | DeepSeek-V4-Flash |
| Version | MiniMax-M3 | DeepSeek-V4-Flash |
| Lifecycle | active | active |
| Released | 2026-06-01 | 2026-04-24 |
| Knowledge cutoff | Unknown | Unknown |
| Input modalities | Text, Image | Text |
| Output modalities | Text | Text |
| Context window | 1,049K | 1,049K |
| Total parameters | 427B | 290.9B |
| Active parameters | 23B | 13B |
| License | other | mit |
| Open weights | Yes | Yes |
| API available | Yes | Yes |
| Self-hostable | Yes | Yes |
| Provider access | Deepinfra (Standard), Fireworks Ai (Standard), Hugging Face (Standard), Openrouter (Standard), Together Ai (Standard) | DeepSeek (Standard), Deepinfra (Standard), Fireworks Ai (Standard), Hugging Face (Standard), Openrouter (Standard) |
| Capabilities | chat, generation, reasoning, tools | chat, generation, reasoning |
17 comparable fields · 11 material differences · Pair passes the primary-source comparison gate
MiniMax M3 Capabilities
DeepSeek V4 Flash Capabilities
Internal Comparison Graph
Related Comparisons
| A | Pair | B | Context |
|---|---|---|---|
MiniMax-M2.7MiniMax | vs | MiniMax-M3MiniMax | family variantstext |
MiniMax-M3MiniMax | vs | DeepSeek-V3.2DeepSeek | cross-developer peerstext |
MiniMax-M3MiniMax | vs |  GPT-6 AstraOpenAI | cross-developer peersimage, text |
MiniMax-M3MiniMax | vs | Claude Fable 5.1Anthropic | cross-developer peersimage, text |
MiniMax-M3MiniMax | vs |  Gemini 3.1 ProGoogle DeepMind | cross-developer peerstext |
MiniMax-M3MiniMax | vs | DeepSeek-V4-ProDeepSeek | cross-developer peerstext |
MiniMax-M3MiniMax | vs | Grok 4.6xAI | cross-developer peersimage, text |
MiniMax-M3MiniMax | vs |  Qwen3.8-MaxQwen | cross-developer peersimage, text |
MiniMax-M3MiniMax | vs |  Kimi-K3Moonshot AI | cross-developer peersimage, text |
MiniMax-M3MiniMax | vs |  GLM-5.3Z.ai | cross-developer peerstext |
MiniMax-M3MiniMax | vs |  Hy4 previewTencent | cross-developer peerstext |
MiniMax-M3MiniMax | vs | Seed 2.1 ProByteDance Seed | cross-developer peersimage, text |
Primary Evidence
Sources and Freshness
Questions
MiniMax M3 vs DeepSeek V4 Flash FAQs
Is MiniMax M3 or DeepSeek V4 Flash better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both MiniMax M3 and DeepSeek V4 Flash, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, MiniMax M3 or DeepSeek V4 Flash?+
MiniMax M3 is $0.28 and DeepSeek V4 Flash is $0.0868 per million tokens, so DeepSeek V4 Flash is cheaper on this metric. MiniMax M3 is $1.10 and DeepSeek V4 Flash is $0.1736 per million tokens, so DeepSeek V4 Flash is cheaper on this metric.
Which has a larger context window, MiniMax M3 or DeepSeek V4 Flash?+
Neither model has a larger sourced context window in this comparison. MiniMax M3 is 1,049K and DeepSeek V4 Flash is 1,049K.
Which performs better in benchmarks, MiniMax M3 or DeepSeek V4 Flash?+
There is no overall benchmark winner: An overall winner requires at least two decisive benchmarks from at least two original publishers.
Can MiniMax M3 or DeepSeek V4 Flash be self-hosted?+
Both models have the same recorded self-hosting status: supported. MiniMax M3 is open weight; DeepSeek V4 Flash is open weight.
Can MiniMax M3 and DeepSeek V4 Flash understand images?+
MiniMax M3 is documented with image input; DeepSeek V4 Flash is not documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, MiniMax M3 or DeepSeek V4 Flash?+
Neither has a larger sourced maximum output. MiniMax M3 is — and DeepSeek V4 Flash is —.
Do MiniMax M3 and DeepSeek V4 Flash support reasoning and tool use?+
MiniMax M3: reasoning, tool calling, and image input. DeepSeek V4 Flash: reasoning. Feature support does not establish relative quality.
Which is available from more inference providers, MiniMax M3 or DeepSeek V4 Flash?+
MiniMax M3 has 5 sourced provider routes; DeepSeek V4 Flash has 5, a tie.
Which offers better value, MiniMax M3 or DeepSeek V4 Flash?+
There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.