Knowledge
Like-for-like- MiniMax M2.7
- 48.7
- Supported · #90/183
- Qwen3.8 Max
- 68.8
- Supported · #17/183
- Basis
- BenchAlign lane · 4 vs 6 public rows
- Reading
- Qwen3.8 Max leads · intervals overlap
MiniMax M2.7 vs Qwen3.8 Max
Decision reading
Qwen3.8 Max has the higher public score, 71.76 versus 55.14, and the 90% score intervals do not overlap.
8 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Updated September 15, 2026. Rank says Qwen3.8 Max is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Share or export
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.
Long documents
Prompts that approach the documented context limit
Qwen3.8 Max
Qwen3.8 Max has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
MiniMax M2.7 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
MiniMax M2.7 is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limited
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
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. MiniMax M2.7 does not fit this workload in one request. MiniMax M2.7 has no published cached-input rate, so cached tokens use its listed input rate. Qwen3.8 Max has no comparable published API token rate.
Confidence: rate-fallback
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.
15 MiniMax M2.7 only
8 Shared
52 Qwen3.8 Max only
- Shared results
- 8
- MiniMax M2.7 only
- 15
- Qwen3.8 Max only
- 52
- Like-for-like categories
- 1 / 8
3 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Agentic
Directional only- MiniMax M2.7
- 41.1
- Estimated · #110/153
- Qwen3.8 Max
- 67.3
- Supported · #9/153
- Basis
- BenchAlign lane · 7 vs 15 public rows
- Reading
- Directional only
Coding
Directional only- MiniMax M2.7
- 48.6
- Estimated · #68/152
- Qwen3.8 Max
- 60.8
- Supported · #20/152
- Basis
- BenchAlign lane · 11 vs 12 public rows
- Reading
- Directional only
Instruction following
Directional only- MiniMax M2.7
- 93.0
- #10/123
- Qwen3.8 Max
- 90.7
- #18/123
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Directional only
Reasoning
Not comparable- MiniMax M2.7
- 74.8
- Unranked · 2 rankable rows
- Qwen3.8 Max
- 86.5
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
Math
Not comparable- MiniMax M2.7
- Not ranked
- Qwen3.8 Max
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- MiniMax M2.7
- Not ranked
- Qwen3.8 Max
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- MiniMax M2.7
- Not ranked
- Qwen3.8 Max
- 87.4
- #5/48
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
| Category | MiniMax M2.7 | Qwen3.8 Max | Basis | Reading |
|---|---|---|---|---|
| Knowledge | 48.7Supported · #90/183 | 68.8Supported · #17/183 | Like-for-likeBenchAlign lane · 4 vs 6 public rows | Qwen3.8 Max leads · intervals overlap |
| Agentic | 41.1Estimated · #110/153 | 67.3Supported · #9/153 | Directional onlyBenchAlign lane · 7 vs 15 public rows | Directional only |
| Coding | 48.6Estimated · #68/152 | 60.8Supported · #20/152 | Directional onlyBenchAlign lane · 11 vs 12 public rows | Directional only |
| Instruction following | 93.0#10/123 | 90.7#18/123 | Directional onlyProvisional lane · 0 vs 1 weighted rows | Directional only |
| Reasoning | 74.8Unranked · 2 rankable rows | 86.5Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
| Math | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | Not ranked | 87.4#5/48 | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
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.
SWE-bench Pro
Coding
MMLU-Pro (Vals)
Knowledge
LiveCodeBench (Vals)
Coding
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
- MiniMax M2.7
- $0.0009
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
Qwen3.8 Max has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- MiniMax M2.7
- $0.0186
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
Qwen3.8 Max has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- MiniMax M2.7
- $0.078
- Does not fit in one request
- Cached input priced at the published list-input rate
- Qwen3.8 Max
- API rate not published
- Fits in one request
- Cached-input rate unavailable
MiniMax M2.7 does not fit this workload in one request. MiniMax M2.7 has no published cached-input rate, so cached tokens use its listed input 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.
SpecificationMiniMax M2.7Qwen3.8 Max
Context window
Maximum documented context; output-token limits may be lower.
MiniMax M2.7
200K
Qwen3.8 Max
API model ID
MiniMax M2.7
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.
MiniMax M2.7
Not published
Qwen3.8 Max
No comparable hosted API rate
Alibaba Cloud Model Studio pricingDocumented inputs
MiniMax M2.7
Not sourced
Qwen3.8 Max
Not sourced
Documented outputs
MiniMax M2.7
Not sourced
Qwen3.8 Max
Not sourced
Provider availability
MiniMax M2.7
Not sourced
Qwen3.8 Max
Not sourced
Reasoning profile
MiniMax M2.7
Non-Reasoning
Qwen3.8 Max
Reasoning
Weight access
MiniMax M2.7
Open Weight
Qwen3.8 Max
Open Weight
License
MiniMax M2.7
Open Weight
Qwen3.8 Max
Open Weight
Release date
MiniMax M2.7
2026-03-18
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
- Qwen3.8 Max has the higher public score, 71.76 versus 55.14, and the 90% score intervals do not overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Qwen3.8 Max has the larger documented window (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 2.0
MiniMax M2.757%SourceQwen3.8 Max—Not directly comparable
Toolathlon
MiniMax M2.746.3%SourceQwen3.8 Max—Not directly comparable
MLE-Bench Lite
MiniMax M2.766.6%SourceQwen3.8 Max—Not directly comparable
MM-ClawBench
MiniMax M2.762.7%SourceQwen3.8 Max—Not directly comparable
Claw-Eval
MiniMax M2.748.7%SourceQwen3.8 Max—Not directly comparable
Gert Labs
MiniMax M2.740.40%SourceQwen3.8 Max—Not directly comparable
Terminal-Bench 2.1 (Vals)
MiniMax M2.748.7%SourceQwen3.8 Max67.4%SourceQwen3.8 Max leads this result
Terminal-Bench 2.1
MiniMax M2.7—Qwen3.8 Max86.6%SourceNot directly comparable
CoWorkBench
MiniMax M2.7—Qwen3.8 Max74.8%SourceNot directly comparable
JobBench
MiniMax M2.7—Qwen3.8 Max53.4%SourceNot directly comparable
skillsBench
MiniMax M2.7—Qwen3.8 Max70.2%SourceNot directly comparable
Agents' Last Exam
MiniMax M2.7—Qwen3.8 Max52.4%SourceNot directly comparable
AutomationBench
MiniMax M2.7—Qwen3.8 Max27.3%SourceNot directly comparable
Toolathlon-Verified
MiniMax M2.7—Qwen3.8 Max72.5%SourceNot directly comparable
WideResearch
MiniMax M2.7—Qwen3.8 Max81.9%SourceNot directly comparable
HLE w/ tools
MiniMax M2.7—Qwen3.8 Max56.2%SourceNot directly comparable
OSWorld-Verified
MiniMax M2.7—Qwen3.8 Max86.1%SourceNot directly comparable
OSWorld 2.0
MiniMax M2.7—Qwen3.8 Max19.4%SourceNot directly comparable
WebArena-Verified
MiniMax M2.7—Qwen3.8 Max66.8%SourceNot directly comparable
AndroidWorld
MiniMax M2.7—Qwen3.8 Max85.3%SourceNot directly comparable
MobileWorld
MiniMax M2.7—Qwen3.8 Max77.8%SourceNot directly comparable
Coding
SWE-bench Verified*
MiniMax M2.775.4%SourceQwen3.8 Max—Not directly comparable
SWE-bench Pro
MiniMax M2.756.2%SourceQwen3.8 Max67.7%SourceQwen3.8 Max leads this result
SWE-Rebench
MiniMax M2.751.9%SourceQwen3.8 Max—Not directly comparable
SWE Multilingual
MiniMax M2.776.5%SourceQwen3.8 Max—Not directly comparable
Multi-SWE Bench
MiniMax M2.752.7%SourceQwen3.8 Max—Not directly comparable
VIBE-Pro
MiniMax M2.755.6%SourceQwen3.8 Max—Not directly comparable
NL2Repo
MiniMax M2.739.8%SourceQwen3.8 Max55.9%SourceQwen3.8 Max leads this result
Vibe Code Bench
MiniMax M2.727.04%SourceQwen3.8 Max—Not directly comparable
React Native Evals
MiniMax M2.771.4%SourceQwen3.8 Max—Not directly comparable
LiveCodeBench (Vals)
MiniMax M2.779.9%SourceQwen3.8 Max87.9%SourceQwen3.8 Max leads this result
SWE-bench (Vals)
MiniMax M2.773.8%SourceQwen3.8 Max85.6%SourceQwen3.8 Max leads this result
Terminal-Bench 2.1
MiniMax M2.7—Qwen3.8 Max86.6%SourceNot directly comparable
DeepSWE
MiniMax M2.7—Qwen3.8 Max56.6%SourceNot directly comparable
FrontierSWE
MiniMax M2.7—Qwen3.8 Max73.5%SourceNot directly comparable
MLS-Bench Lite
MiniMax M2.7—Qwen3.8 Max41.0%SourceNot directly comparable
PaperBench
MiniMax M2.7—Qwen3.8 Max93.0%SourceNot directly comparable
VulcanBench v3
MiniMax M2.7—Qwen3.8 Max81.2%SourceNot directly comparable
OpenHarmony Bench
MiniMax M2.7—Qwen3.8 Max60.8%SourceNot directly comparable
FrontierSWE v2
MiniMax M2.7—Qwen3.8 Max15.8%SourceNot directly comparable
Reasoning
Knowledge
GPQA-D
MiniMax M2.787.0%SourceQwen3.8 Max92.6%SourceQwen3.8 Max leads this result
MMLU-Pro (Arcee)
MiniMax M2.780.8%SourceQwen3.8 Max—Not directly comparable
GPQA Diamond (Vals)
MiniMax M2.786.6%SourceQwen3.8 Max93.7%SourceQwen3.8 Max leads this result
MMLU-Pro (Vals)
MiniMax M2.780.4%SourceQwen3.8 Max88.6%SourceQwen3.8 Max leads this result
GPQA
MiniMax M2.7—Qwen3.8 Max92.6%SourceNot directly comparable
HLE
MiniMax M2.7—Qwen3.8 Max43.6%SourceNot directly comparable
HLE w/o tools
MiniMax M2.7—Qwen3.8 Max43.6%SourceNot directly comparable
Math
AIME25 (Arcee)
MiniMax M2.780.0%SourceQwen3.8 Max—Not directly comparable
Multimodal
MMMU-Pro
MiniMax M2.7—Qwen3.8 Max82.3%SourceNot directly comparable
MathVision
MiniMax M2.7—Qwen3.8 Max95.2%SourceNot directly comparable
MathVision w/ Python
MiniMax M2.7—Qwen3.8 Max97.7%SourceNot directly comparable
BabyVision
MiniMax M2.7—Qwen3.8 Max82.0%SourceNot directly comparable
BabyVision w/ Python
MiniMax M2.7—Qwen3.8 Max91.3%SourceNot directly comparable
ZeroBench
MiniMax M2.7—Qwen3.8 Max24.0%SourceNot directly comparable
ZeroBench w/ Python
MiniMax M2.7—Qwen3.8 Max49.0%SourceNot directly comparable
MedXpertQA (MM)
MiniMax M2.7—Qwen3.8 Max80.4%SourceNot directly comparable
ScreenSpot Pro
MiniMax M2.7—Qwen3.8 Max84.5%SourceNot directly comparable
Vision2Web
MiniMax M2.7—Qwen3.8 Max69.0%SourceNot directly comparable
CharXiv w/o tools
MiniMax M2.7—Qwen3.8 Max88.4%SourceNot directly comparable
CharXiv
MiniMax M2.7—Qwen3.8 Max93.5%SourceNot directly comparable
OmniDocBench 1.5
MiniMax M2.7—Qwen3.8 Max92.1%SourceNot directly comparable
OCRBench V2
MiniMax M2.7—Qwen3.8 Max74.2%SourceNot directly comparable
CC-OCR
MiniMax M2.7—Qwen3.8 Max79.6%SourceNot directly comparable
RealWorldQA
MiniMax M2.7—Qwen3.8 Max88.0%SourceNot directly comparable
ERQA
MiniMax M2.7—Qwen3.8 Max77.8%SourceNot directly comparable
SimpleVQA
MiniMax M2.7—Qwen3.8 Max75.0%SourceNot directly comparable
PerceptionBench
MiniMax M2.7—Qwen3.8 Max63.5%SourceNot directly comparable
Video-MME (with subtitle)
MiniMax M2.7—Qwen3.8 Max90.4%SourceNot directly comparable
VideoMMMU
MiniMax M2.7—Qwen3.8 Max88.7%SourceNot directly comparable
MMVU
MiniMax M2.7—Qwen3.8 Max82.4%SourceNot directly comparable
MLVU (M-Avg)
MiniMax M2.7—Qwen3.8 Max90.8%SourceNot directly comparable
LVBench
MiniMax M2.7—Qwen3.8 Max81.8%SourceNot directly comparable
Instruction following
IFBench
MiniMax M2.7—Qwen3.8 Max82.8%SourceNot directly comparable
Questions
Which is better, MiniMax M2.7 or Qwen3.8 Max?
Qwen3.8 Max has the higher public score, 71.76 versus 55.14, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.
Which is better for coding, MiniMax M2.7 or Qwen3.8 Max?
Qwen3.8 Max scores higher for coding on the public lane, 60.8 to 48.6. MiniMax M2.7 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, MiniMax M2.7 or Qwen3.8 Max?
Qwen3.8 Max scores higher for agentic tasks on the public lane, 67.3 to 41.1. MiniMax M2.7 is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which costs less, MiniMax M2.7 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, MiniMax M2.7 or Qwen3.8 Max?
Qwen3.8 Max has the larger documented context window: 1M, compared with 200K.
Related comparisons
Last updated September 15, 2026
Watch MiniMax M2.7 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.