Reasoning
Directional only- DeepSeek V4 Pro Base
- 51.5
- Qwen3.8 Max
- 78.3
- Weighted basis
- 1 vs 2 rows
- Reading
- Directional only
Model comparison
DeepSeek V4 Pro Base vs Qwen3.8 Max
Updated August 3, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
1 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
No workload recommendation clears the current evidence threshold.
Use the matched evidence, workload costs, and sourced specifications below instead of treating a point score as a universal answer.
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Long documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
23 DeepSeek V4 Pro Base only
1 Shared
51 Qwen3.8 Max only
- Shared results
- 1
- DeepSeek V4 Pro Base only
- 23
- Qwen3.8 Max only
- 51
- Like-for-like categories
- 0 / 8
1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Agentic
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- 86.1
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Coding
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- 67.7
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Knowledge
Not comparable- DeepSeek V4 Pro Base
- 66.4
- Qwen3.8 Max
- 50.2
- Weighted basis
- 3 vs 2 rows
- Reading
- Not comparable
Math
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- 86.3
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- DeepSeek V4 Pro Base
- Not measured
- Qwen3.8 Max
- 82.8
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
| Category | DeepSeek V4 Pro Base | Qwen3.8 Max | Weighted basis | Reading |
|---|---|---|---|---|
| Reasoning | 51.5 | 78.3 | Directional only1 vs 2 rows | Directional only |
| Agentic | Not measured | 86.1 | Not comparable0 vs 1 rows | Not comparable |
| Coding | Not measured | 67.7 | Not comparable0 vs 1 rows | Not comparable |
| Knowledge | 66.4 | 50.2 | Not comparable3 vs 2 rows | Not comparable |
| Math | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 86.3 | Not comparable0 vs 2 rows | Not comparable |
| Instruction following | Not measured | 82.8 | Not comparable0 vs 1 rows | Not comparable |
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
LongBench v2
Reasoning
DeepSeek V4 Pro Base: 51.5%Qwen3.8 Max: 66.3%Normalized gap 14.8DeepSeek V4 Pro Base source Qwen3.8 Max source
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- DeepSeek V4 Pro Base
- API rate not published
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
DeepSeek V4 Pro Base has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- DeepSeek V4 Pro Base
- API rate not published
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
DeepSeek V4 Pro Base has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- DeepSeek V4 Pro Base
- API rate not published
- Fits in one request
- Cached-input rate unavailable
- Qwen3.8 Max
- API rate not published
- Fits in one request
- Cached-input rate unavailable
DeepSeek V4 Pro Base has no comparable published API token rate. Qwen3.8 Max has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationDeepSeek V4 Pro BaseQwen3.8 Max
Context window
Maximum documented context; output-token limits may be lower.
DeepSeek V4 Pro Base
Qwen3.8 Max
API model ID
DeepSeek V4 Pro Base
Not sourced
Qwen3.8 Max
qwen3.8-max
Alibaba Cloud Model Studio pricingCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
DeepSeek V4 Pro Base
No comparable hosted API rate
Qwen3.8 Max
No comparable hosted API rate
Alibaba Cloud Model Studio pricingDocumented inputs
DeepSeek V4 Pro Base
Qwen3.8 Max
Not sourced
Documented outputs
DeepSeek V4 Pro Base
Qwen3.8 Max
Not sourced
Provider availability
DeepSeek V4 Pro Base
Open Weights · open weights
DeepSeek V4 Pro model cardQwen3.8 Max
Not sourced
Reasoning profile
DeepSeek V4 Pro Base
Non-Reasoning
Qwen3.8 Max
Reasoning
Weight access
DeepSeek V4 Pro Base
Open Weight
Qwen3.8 Max
Proprietary
License
DeepSeek V4 Pro Base
Open Weight
Qwen3.8 Max
Proprietary
Release date
DeepSeek V4 Pro Base
2026-04-24
Qwen3.8 Max
2026-08-03
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Both models list 1M.
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence75 rows
Agentic
Terminal-Bench 2.1
DeepSeek V4 Pro Base—Qwen3.8 Max86.6%SourceNot directly comparable
CoWorkBench
DeepSeek V4 Pro Base—Qwen3.8 Max74.8%SourceNot directly comparable
JobBench
DeepSeek V4 Pro Base—Qwen3.8 Max53.4%SourceNot directly comparable
skillsBench
DeepSeek V4 Pro Base—Qwen3.8 Max70.2%SourceNot directly comparable
Agents' Last Exam
DeepSeek V4 Pro Base—Qwen3.8 Max52.4%SourceNot directly comparable
AutomationBench
DeepSeek V4 Pro Base—Qwen3.8 Max27.3%SourceNot directly comparable
Toolathlon-Verified
DeepSeek V4 Pro Base—Qwen3.8 Max72.5%SourceNot directly comparable
WideResearch
DeepSeek V4 Pro Base—Qwen3.8 Max81.9%SourceNot directly comparable
HLE w/ tools
DeepSeek V4 Pro Base—Qwen3.8 Max56.2%SourceNot directly comparable
OSWorld-Verified
DeepSeek V4 Pro Base—Qwen3.8 Max86.1%SourceNot directly comparable
OSWorld 2.0
DeepSeek V4 Pro Base—Qwen3.8 Max19.4%SourceNot directly comparable
WebArena-Verified
DeepSeek V4 Pro Base—Qwen3.8 Max66.8%SourceNot directly comparable
AndroidWorld
DeepSeek V4 Pro Base—Qwen3.8 Max85.3%SourceNot directly comparable
MobileWorld
DeepSeek V4 Pro Base—Qwen3.8 Max77.8%SourceNot directly comparable
Coding
BigCodeBench
DeepSeek V4 Pro Base59.2%SourceQwen3.8 Max—Not directly comparable
HumanEval
DeepSeek V4 Pro Base76.8%SourceQwen3.8 Max—Not directly comparable
Terminal-Bench 2.1
DeepSeek V4 Pro Base—Qwen3.8 Max86.6%SourceNot directly comparable
SWE-bench Pro
DeepSeek V4 Pro Base—Qwen3.8 Max67.7%SourceNot directly comparable
deepSwe
DeepSeek V4 Pro Base—Qwen3.8 Max56.6%SourceNot directly comparable
NL2Repo
DeepSeek V4 Pro Base—Qwen3.8 Max55.9%SourceNot directly comparable
FrontierSWE
DeepSeek V4 Pro Base—Qwen3.8 Max73.5%SourceNot directly comparable
MLS-Bench Lite
DeepSeek V4 Pro Base—Qwen3.8 Max41.0%SourceNot directly comparable
PaperBench
DeepSeek V4 Pro Base—Qwen3.8 Max93.0%SourceNot directly comparable
Reasoning
BBH
DeepSeek V4 Pro Base87.5%SourceQwen3.8 Max—Not directly comparable
DROP
DeepSeek V4 Pro Base88.7%SourceQwen3.8 Max—Not directly comparable
HellaSwag
DeepSeek V4 Pro Base88.0%SourceQwen3.8 Max—Not directly comparable
WinoGrande
DeepSeek V4 Pro Base81.5%SourceQwen3.8 Max—Not directly comparable
CLUEWSC
DeepSeek V4 Pro Base85.2%SourceQwen3.8 Max—Not directly comparable
LongBench v2
DeepSeek V4 Pro Base51.5%SourceQwen3.8 Max66.3%SourceQwen3.8 Max leads this result
MRCRv2
DeepSeek V4 Pro Base—Qwen3.8 Max92.9%SourceNot directly comparable
Knowledge
AGIEval
DeepSeek V4 Pro Base83.1%SourceQwen3.8 Max—Not directly comparable
MMLU
DeepSeek V4 Pro Base90.1%SourceQwen3.8 Max—Not directly comparable
MMLU-Redux
DeepSeek V4 Pro Base90.8%SourceQwen3.8 Max—Not directly comparable
MMLU-Pro
DeepSeek V4 Pro Base73.5%SourceQwen3.8 Max—Not directly comparable
MMMLU
DeepSeek V4 Pro Base90.3%SourceQwen3.8 Max—Not directly comparable
C-Eval
DeepSeek V4 Pro Base93.1%SourceQwen3.8 Max—Not directly comparable
CMMLU
DeepSeek V4 Pro Base90.8%SourceQwen3.8 Max—Not directly comparable
MultiLoKo
DeepSeek V4 Pro Base51.1%SourceQwen3.8 Max—Not directly comparable
SimpleQA
DeepSeek V4 Pro Base55.2%SourceQwen3.8 Max—Not directly comparable
SuperGPQA
DeepSeek V4 Pro Base53.9%SourceQwen3.8 Max—Not directly comparable
FACTS Parametric
DeepSeek V4 Pro Base62.6%SourceQwen3.8 Max—Not directly comparable
TriviaQA
DeepSeek V4 Pro Base85.6%SourceQwen3.8 Max—Not directly comparable
GPQA
DeepSeek V4 Pro Base—Qwen3.8 Max92.6%SourceNot directly comparable
GPQA-D
DeepSeek V4 Pro Base—Qwen3.8 Max92.6%SourceNot directly comparable
HLE
DeepSeek V4 Pro Base—Qwen3.8 Max43.6%SourceNot directly comparable
HLE w/o tools
DeepSeek V4 Pro Base—Qwen3.8 Max43.6%SourceNot directly comparable
Math
Multilingual
MGSM
DeepSeek V4 Pro Base84.4%SourceQwen3.8 Max—Not directly comparable
Multimodal
MMMU-Pro
DeepSeek V4 Pro Base—Qwen3.8 Max82.3%SourceNot directly comparable
MathVision
DeepSeek V4 Pro Base—Qwen3.8 Max95.2%SourceNot directly comparable
MathVision w/ Python
DeepSeek V4 Pro Base—Qwen3.8 Max97.7%SourceNot directly comparable
BabyVision
DeepSeek V4 Pro Base—Qwen3.8 Max82.0%SourceNot directly comparable
BabyVision w/ Python
DeepSeek V4 Pro Base—Qwen3.8 Max91.3%SourceNot directly comparable
ZeroBench
DeepSeek V4 Pro Base—Qwen3.8 Max24.0%SourceNot directly comparable
ZeroBench w/ Python
DeepSeek V4 Pro Base—Qwen3.8 Max49.0%SourceNot directly comparable
MedXpertQA (MM)
DeepSeek V4 Pro Base—Qwen3.8 Max80.4%SourceNot directly comparable
ScreenSpot Pro
DeepSeek V4 Pro Base—Qwen3.8 Max84.5%SourceNot directly comparable
Vision2Web
DeepSeek V4 Pro Base—Qwen3.8 Max69.0%SourceNot directly comparable
CharXiv w/o tools
DeepSeek V4 Pro Base—Qwen3.8 Max88.4%SourceNot directly comparable
CharXiv
DeepSeek V4 Pro Base—Qwen3.8 Max93.5%SourceNot directly comparable
OmniDocBench 1.5
DeepSeek V4 Pro Base—Qwen3.8 Max92.1%SourceNot directly comparable
OCRBench V2
DeepSeek V4 Pro Base—Qwen3.8 Max74.2%SourceNot directly comparable
CC-OCR
DeepSeek V4 Pro Base—Qwen3.8 Max79.6%SourceNot directly comparable
RealWorldQA
DeepSeek V4 Pro Base—Qwen3.8 Max88.0%SourceNot directly comparable
ERQA
DeepSeek V4 Pro Base—Qwen3.8 Max77.8%SourceNot directly comparable
SimpleVQA
DeepSeek V4 Pro Base—Qwen3.8 Max75.0%SourceNot directly comparable
PerceptionBench
DeepSeek V4 Pro Base—Qwen3.8 Max63.5%SourceNot directly comparable
Video-MME (with subtitle)
DeepSeek V4 Pro Base—Qwen3.8 Max90.4%SourceNot directly comparable
VideoMMMU
DeepSeek V4 Pro Base—Qwen3.8 Max88.7%SourceNot directly comparable
MMVU
DeepSeek V4 Pro Base—Qwen3.8 Max82.4%SourceNot directly comparable
MLVU (M-Avg)
DeepSeek V4 Pro Base—Qwen3.8 Max90.8%SourceNot directly comparable
LVBench
DeepSeek V4 Pro Base—Qwen3.8 Max81.8%SourceNot directly comparable
Instruction following
IFBench
DeepSeek V4 Pro Base—Qwen3.8 Max82.8%SourceNot directly comparable
Frequently asked questions
Which is better, DeepSeek V4 Pro Base or Qwen3.8 Max?
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, DeepSeek V4 Pro Base or Qwen3.8 Max?
The published evidence does not provide a shared weighted coding basis for both models, so BenchLM does not name a coding winner.
Which is better for agentic tasks, DeepSeek V4 Pro Base or Qwen3.8 Max?
The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.
Which costs less, DeepSeek V4 Pro Base or Qwen3.8 Max?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, DeepSeek V4 Pro Base or Qwen3.8 Max?
Both models list the same context window, 1M.
Related comparisons
Last updated August 3, 2026
Watch DeepSeek V4 Pro Base vs Qwen3.8 Max
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.