DeepSeek V4 Flash vs Qwen4 Max
Full profiles: DeepSeek V4 Flash (DeepSeek) and Qwen4 Max (Alibaba): every benchmark, price and rank.
Updated · Data from 10 sources
As of 19 Nov 2026, there are not enough shared benchmarks for a verdict, and DeepSeek V4 Flash is 8.6× cheaper at $0.262 vs $2.25 per 1M tokens blended. Qwen4 Max leads in every category we measure both in. DeepSeek V4 Flash streams ~1.5× faster.
Key facts (as of 19 Nov 2026)
- DeepSeek V4 Flash costs $0.262 per 1M tokens blended (list price, 19 Nov 2026).
- Qwen4 Max costs $2.25 per 1M tokens blended (list price, 19 Nov 2026).
- DeepSeek V4 Flash streams ~1.5× faster than Qwen4 Max (Artificial Analysis, 7-day median, 19 Nov 2026).
Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | Either | 55 vs 55 category score | Either |
| Reasoning | Either | 59 vs 59 category score | Either |
| Knowledge | Qwen4 Max | 59 vs 55 category score | Medium |
| Chat quality (human preference) | – | Only DeepSeek V4 Flash has an Arena rating (62 category score) | Not comparable |
| Lowest cost | DeepSeek V4 Flash | $0.262 vs $2.25 per 1M blended; agentic coding session $0.011 vs $0.108 | – |
| Fastest first answer | DeepSeek V4 Flash | 0.6 s (DeepSeek V4 Flash (high)) vs 139.2 s (Qwen4 Max) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Not enough shared benchmarks for a verdict.
- DeepSeek V4 Flash (high) is stronger in reasoning.
- Qwen4 Max is stronger in coding, knowledge.
- DeepSeek V4 Flash (high) is 8.6× cheaper ($0.262 vs $2.25 per 1M blended).
- DeepSeek V4 Flash (high) streams 1.5× faster (86 vs 59 tokens per second).
At a glance
| Spec | DeepSeek V4 Flash | Qwen4 Max |
|---|---|---|
| Developer | DeepSeek | Alibaba |
| Released | 31 Jul 2026 | 16 Nov 2026 |
| Context window (tokens) | 1,048,576 | 991,808 |
| Compared at | DeepSeek V4 Flash (high) | Qwen4 Max |
| Open weights | Yes | No |
| Price in / out / cached input per 1M | $0.150 / $0.600 / $0.0030 | $1.50 / $4.50 / $0.071 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | DeepSeek V4 Flash (high) | Qwen4 Max | Leader |
|---|---|---|---|
| Reasoning | 58.7 | 58.5 | level |
| Coding | 54.8 | 55.1 | level |
| Maths | 61.8 | – | not comparable |
| Knowledge | 54.8 | 59.0 | Qwen4 Max |
| Human preference | 62.0 | – | not comparable |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Not drawn, measured for fewer than two of these models: Agents & tools, Maths, Instruction following, Long context, Multimodal, Human preference.
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | DeepSeek V4 Flash (high) | Qwen4 Max |
|---|---|---|
| Blended price per 1M tokens | $0.262 | $2.25 |
| Input / output per 1M | $0.150 / $0.600 | $1.50 / $4.50 |
| Cached input per 1M | $0.0030 | $0.071 |
| Price per task: chat reply | $0.0002 | $0.0020 |
| Price per task: summarize a 30-page report | $0.0022 | $0.021 |
| Price per task: code edit | $0.0018 | $0.016 |
| Price per task: agentic coding session | $0.011 | $0.108 |
| Price per task: structured extraction | $0.0004 | $0.0039 |
| Agentic coding session, input mostly cached | $0.0048 | $0.044 |
| Cost per benchmark task (AA Index) | – | $3.27 |
| Output tokens per task (AA Index) | – | 105k |
| Of which reasoning | – | 69k |
| Time per task (AA Index) | – | 30 min |
| Output speed | 86 tok/s | 59 tok/s |
| Time to first answer | 0.6 s | 139.2 s |
Each model is compared at its best-ranked reasoning-effort setting. DeepSeek V4 Flash (high) scores 57.0 and takes 1 s to its first answer token; DeepSeek V4 Flash (max) scores 56.7 in 21 s, DeepSeek V4 Flash scores 53.8 in 1 s. Qwen4 Max and takes 139 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
113 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | DeepSeek V4 Flash (high) | Qwen4 Max |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 57.0 | – |
| Coding score | BenchLeader | 54.8 | 55.1 |
| Human preference score | BenchLeader | 62.0 | – |
| Knowledge score | BenchLeader | 54.8 | 59.0 |
| Maths score | BenchLeader | 61.8 | – |
| Reasoning score | BenchLeader | 58.7 | 58.5 |
| Blended price $/M | BenchLeader | $0.262 | $2.25 |
| Output speed | BenchLeader | 86 tok/s | 59 tok/s |
| Time to first answer | BenchLeader | 0.6 s | 139.2 s |
| Context window | BenchLeader | 1,048,576 tokens | 991,808 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 91.0%† | – |
| SimpleBench | SimpleBench | 61.1%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1459 | – |
| LiveBench 2.0 Reasoning | LiveBench | – | 92.2%† |
| LiveBench 1.0 Reasoning | LiveBench | 86.6%† | – |
| GPQA Diamond (AA) | Artificial Analysis | 90.8%† | – |
| Humanity's Last Exam (AA) | Artificial Analysis | 38.5%† | – |
| HLE-Diamond (AA) | Artificial Analysis | – | 44.9% |
| GPQA Diamond (Vals) | Vals AI | 89.9% | – |
| Kagi LLM Benchmark | Kagi LLM Benchmark | 52.2%† | – |
| ARC-AGI-1 | ARC Prize | 87.0% | – |
| ARC-AGI-2 | ARC Prize | 56.0% | – |
| CritPt | Artificial Analysis | 16.6%† | – |
| CritPt v2 hard subset (AA) | Artificial Analysis | – | 43.0% |
| Arena Expert | Arena (LMArena) | 1467 | – |
| Chess Puzzles | Epoch AI Benchmarking Hub | 32.0%† | – |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 34.0%† | – |
| LMCA | Epoch AI Benchmarking Hub | 41.7%† | – |
| DTBench | Epoch AI Benchmarking Hub | 90.9%† | – |
| Coding | |||
| SciCode | SciCode | 42.0% | – |
| WeirdML | WeirdML | 57.0% | – |
| FrontierCode | Cognition | 18.8%† | – |
| Arena Coding | Arena (LMArena) | 1480 | – |
| Arena WebDev | Arena (LMArena) | 1582 | 1750† |
| LiveBench 2.0 Coding | LiveBench | – | 69.8%† |
Show all 113 benchmarks
| Metric | Publisher | DeepSeek V4 Flash (high) | Qwen4 Max |
|---|---|---|---|
| Coding | |||
| LiveBench 1.0 Coding | LiveBench | 75.0%† | – |
| SciCode (AA) | Artificial Analysis | 50.4%† | – |
| LiveCodeBench | Vals AI | 87.3% | – |
| IOI | Vals AI | 32.7% | – |
| SWE-bench (Vals) | Vals AI | 88.8% | – |
| Code Migration | Vals AI | 38.6% | – |
| ProgramBench | Vals AI | 0.0% | – |
| Vibe Code Bench v1.1 | Vals AI | 74.7% | 86.7%† |
| ALE-Bench | Epoch AI Benchmarking Hub | 678.2 | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 12.1%† | 48.5% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 78.7%† | – |
| GameBench 2 Overall | DuelLab GameBench 2 | 34.1† | – |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | 43.4 | – |
| FrontierCode 2.0 (AA) | Artificial Analysis | – | 26.4% |
| Agents & tools | |||
| Arena Agent | Arena (LMArena) | – | 7.3† |
| LiveBench Long Agentic | LiveBench | – | 46.6%† |
| LiveBench Browser | LiveBench | – | 80.7%† |
| LiveBench Agentic Coding | LiveBench | 46.8%† | – |
| Terminal-Bench 2.1 (Vals) | Vals AI | 67.0%† | – |
| HiL-Bench | Scale AI SEAL | 47.8%† | – |
| GDPval-AA v2.1 | Artificial Analysis | 1437† | 1809 |
| τ³-Banking (AA) | Artificial Analysis | 39.4%† | – |
| Analyst Agent (AA) | Artificial Analysis | – | 46.3% |
| APEX-Agents (AA) | Artificial Analysis | – | 59.1% |
| CyberBench | Vals AI | 72.1% | – |
| Finance Agent v2 | Vals AI | 49.5% | – |
| Finance Agent v2.1 | Vals AI | – | 66.7%† |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | – | 21.3%† |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 8.3% | – |
| Legal Research Bench | Vals AI | 30.3% | 41.4%† |
| SkillsBench | Vals AI | 50.7% | – |
| Tax Agent Bench | Vals AI | 28.1% | 40.4%† |
| Terminal-Bench 4.0 (Vals) | Vals AI | 18.7% | 48.5%† |
| Terminal-Bench Science | Vals AI | 4.3% | – |
| Arena Agent Task Completion | Arena (LMArena) | – | 8.8† |
| AutomationBench | Artificial Analysis | 54.0%† | 72.6% |
| GDP.pdf (AA) | Artificial Analysis | 11.0%† | 26.2% |
| Harvey LAB-AA v1.1 | Harvey | – | 9.4% |
| EnterpriseOps-Gym | Artificial Analysis | – | 48.5% |
| AA-Briefcase v1.1 | Artificial Analysis | 1255† | 1686 |
| Design Arena Agentic | Design Arena | – | 1356 |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | – | 23.3% |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 57.5%† | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 24.4%† | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 94.4%† | – |
| ProofBench | Vals AI | 56.0%† | – |
| LiveBench Mathematics | LiveBench | 86.8%† | – |
| AIME 2026 | MathArena | 95.8%† | – |
| HMMT February 2026 | MathArena | 93.9%† | – |
| MathArena Apex | MathArena | 27.1%† | – |
| Arena Maths | Arena (LMArena) | 1441 | – |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 33.6%† | – |
| LiveBench 2.0 Data Analysis | LiveBench | – | 83.4%† |
| LiveBench Research | LiveBench | – | 84.7%† |
| LiveBench 1.0 Data Analysis | LiveBench | 79.3%† | – |
| AA-Omniscience v1.1 | Artificial Analysis | – | 27.7 |
| MMLU-Pro | Vals AI | 86.2% | – |
| LegalBench | Vals AI | 77.7% | – |
| CorpFin | Vals AI | 61.9% | – |
| TaxEval | Vals AI | 70.7% | – |
| BioMysteryBench | Vals AI | 64.4% | – |
| Excel Modeling Benchmark | Vals AI | 57.0% | – |
| Excel Modeling Benchmark v1.1 | Vals AI | – | 70.8%† |
| MedCode | Vals AI | 41.4% | – |
| MedScribe | Vals AI | 80.4% | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | – | 43.7% |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | – | 71.6% |
| Instruction following | |||
| LiveBench Language | LiveBench | 79.2%† | – |
| LiveBench Instruction Following | LiveBench | 65.5%† | – |
| Arena Instruction Following | Arena (LMArena) | 1434 | – |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1439 | – |
| EQ-Bench Creative Writing v3 | EQ-Bench | 1559† | – |
| Arena Creative Writing | Arena (LMArena) | 1411 | – |
| Arena Multi-turn | Arena (LMArena) | 1447 | – |
| Arena Longer Queries | Arena (LMArena) | 1449 | – |
| Multimodal | |||
| LiveBench Design | LiveBench | – | 79.1%† |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 79.7%† | – |
| MLCR | Artificial Analysis | – | 19.4% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 154.4† | – |
| LiveBench 2.0 | LiveBench | – | 73.1%† |
| LiveBench 1.0 | LiveBench | 74.2%† | – |
| AA Intelligence Index v5.0.1 | Artificial Analysis | – | 42.9 |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 34.3† | – |
| AA Progress Index v1 | Artificial Analysis | 309.4† | – |
| Vals Index | Vals AI | – | 63.2† |
| Vals Finance Index | Vals AI | – | 62.0%† |
| Vals Legal Index | Vals AI | – | 70.7%† |
| Metadata | |||
| AA Openness Index | Artificial Analysis | 44.4† | – |
Data as of 2026-11-19. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261114-ca113f475f5e (https://benchmarks.duellab.org/releases/gb2-20261114-ca113f475f5e/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- DeepSeek V4 Flash (high): 1 score revision.
- DeepSeek V4 Flash (high): 1 score revision.
- DeepSeek V4 Flash (high): 1 score revision.
- DeepSeek V4 Flash (high): 1 new result (GameBench 2 by reasoning setting).
- DeepSeek V4 Flash (high): 1 new result (Harvey's Legal Agent Benchmark v1).
- DeepSeek V4 Flash (high): BenchLeader Index 59.3 → 57.3.
- DeepSeek V4 Flash (high): input price moved from $0.11/M to $0.15/M (36% up); output price moved from $0.23/M to $0.60/M (161% up).
Related comparisons
- DeepSeek V4 Flash vs Claude Fable 5.1
- DeepSeek V4 Flash vs GPT-6.1 Sol
- Qwen4 Max vs Qwen3.8 Max
- Qwen4 Max vs Inkling 2 Small
- DeepSeek V4 Flash vs Qwen3.8 Max
- DeepSeek V4 Flash vs Gemini 4 Argon
- DeepSeek V4 Flash vs Claude Sonnet 5.5
- DeepSeek V4 Flash vs Muse Spark 1.3
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
DeepSeek V4 Flash vs Qwen4 Max: questions
- DeepSeek V4 Flash vs Qwen4 Max: which is cheaper, and will that change?
- DeepSeek V4 Flash is cheaper today: $0.262 against $2.25 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is DeepSeek V4 Flash or Qwen4 Max better for coding?
- They are level in coding (55 vs 55 on the category index).
- Which is faster, DeepSeek V4 Flash or Qwen4 Max?
- DeepSeek V4 Flash streams ~1.5× faster (86 vs 59 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- DeepSeek V4 Flash accepts more context: 1,048,576 against 991,808 tokens for Qwen4 Max.
- Which has lower latency?
- DeepSeek V4 Flash answers first: 0.6 s against 139.2 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
More questions (1)
- When was Qwen4 Max released?
- Qwen4 Max was released on 16 Nov 2026.