Agentic work
Tool use, computer use, and multi-step task completion
Claude Sonnet 5.5
Claude Sonnet 5.5 leads on the public agentic lane, 66.1 to 64.8, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
Claude Sonnet 5.5 vs Qwen3.8 Max
Updated September 28, 2026. Rank says Claude Sonnet 5.5 is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Share or export
Decision reading
Claude Sonnet 5.5 has the higher public score estimate, 80.49 versus 71.69, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 6 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 6
- Claude Sonnet 5.5 only
- 41
- Qwen3.8 Max only
- 54
- Like-for-like categories
- 2 / 8
Estimated: Claude Sonnet 5.5 · Supported: Qwen3.8 MaxHow the comparison works
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.
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Claude Sonnet 5.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limitedLong documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documentedChat 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-ratesCache-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-ratesRepository 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
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
79.6Claude Sonnet 5.555.6Qwen3.8 Max
Directional only · BenchAlign v5.7
Claude Sonnet 5.5 scores higher, but at least one score rests on Estimated evidence or a different benchmark set. Directional only, no winner.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
41 Claude Sonnet 5.5 only
6 Shared
54 Qwen3.8 Max only
1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
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 ProCoding
Normalized gap 13.6Claude Sonnet 5.5:81.3%Qwen3.8 Max:67.7%HLE w/o toolsKnowledge
Normalized gap 13.3Claude Sonnet 5.5:56.9%Qwen3.8 Max:43.6%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.7 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
Like-for-like- Claude Sonnet 5.5
- 66.1
- Supported · #10/111
- Qwen3.8 Max
- 64.8
- Supported · #11/111
- Basis
- BenchAlign v5.7 lane · 11 vs 15 public rows
- Reading
- Claude Sonnet 5.5 leads · intervals overlap
Knowledge
Like-for-like- Claude Sonnet 5.5
- 80.9
- Supported · #6/160
- Qwen3.8 Max
- 66.6
- Supported · #19/160
- Basis
- BenchAlign v5.7 lane · 16 vs 6 public rows
- Reading
- Claude Sonnet 5.5 leads · intervals overlap
Coding
Directional only- Claude Sonnet 5.5
- 79.6
- Estimated · #3/136
- Qwen3.8 Max
- 55.6
- Supported · #24/136
- Basis
- BenchAlign v5.7 lane · 9 vs 12 public rows
- Reading
- Directional only
Reasoning
Not comparable- Claude Sonnet 5.5
- 79.2
- #7/27
- Qwen3.8 Max
- 87.7
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Sonnet 5.5
- 83.3
- Unranked · 7 rankable rows
- Qwen3.8 Max
- 88.4
- #5/50
- Basis
- Provisional lane · 1 vs 2 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Sonnet 5.5
- Not ranked
- Qwen3.8 Max
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Sonnet 5.5
- Not ranked
- Qwen3.8 Max
- 90.5
- #16/124
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Math
Not comparable- Claude Sonnet 5.5
- Not ranked
- Qwen3.8 Max
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
| Category | Claude Sonnet 5.5 | Qwen3.8 Max | Basis | Reading |
|---|---|---|---|---|
| Agentic | 66.1Supported · #10/111 | 64.8Supported · #11/111 | Like-for-likeBenchAlign v5.7 lane · 11 vs 15 public rows | Claude Sonnet 5.5 leads · intervals overlap |
| Knowledge | 80.9Supported · #6/160 | 66.6Supported · #19/160 | Like-for-likeBenchAlign v5.7 lane · 16 vs 6 public rows | Claude Sonnet 5.5 leads · intervals overlap |
| Coding | 79.6Estimated · #3/136 | 55.6Supported · #24/136 | Directional onlyBenchAlign v5.7 lane · 9 vs 12 public rows | Directional only |
| Reasoning | 79.2#7/27 | 87.7Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
| Multimodal | 83.3Unranked · 7 rankable rows | 88.4#5/50 | Not comparableProvisional lane · 1 vs 2 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | 90.5#16/124 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Math | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
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
- Claude Sonnet 5.5
- $0.007
- 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
- Claude Sonnet 5.5
- $0.13
- 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
- Claude Sonnet 5.5
- $0.18
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
- Cached-input rate unavailable
Qwen3.8 Max has no comparable published API token rate.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationClaude Sonnet 5.5Qwen3.8 Max
Context window
Maximum documented context; output-token limits may be lower.
Claude Sonnet 5.5
Qwen3.8 Max
API model ID
Claude Sonnet 5.5
claude-sonnet-5-5
Anthropic Claude Sonnet 5.5 model documentationQwen3.8 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.
Claude Sonnet 5.5
$0.2 per 1M cached input tokens
Claude API pricingQwen3.8 Max
No comparable hosted API rate
Alibaba Cloud Model Studio pricingDocumented inputs
Claude Sonnet 5.5
Qwen3.8 Max
Not sourced
Documented outputs
Claude Sonnet 5.5
Qwen3.8 Max
Not sourced
Provider availability
Claude Sonnet 5.5
Generally Available · Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Platform on AWS
Anthropic Claude Sonnet 5.5 model documentationQwen3.8 Max
Not sourced
Reasoning profile
Claude Sonnet 5.5
Reasoning
Qwen3.8 Max
Reasoning
Weight access
Claude Sonnet 5.5
Proprietary
Qwen3.8 Max
Open Weight
License
Claude Sonnet 5.5
Proprietary
Qwen3.8 Max
Open Weight
Release date
Claude Sonnet 5.5
2026-09-28
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
- Claude Sonnet 5.5 has the higher public score estimate, 80.49 versus 71.69, but the 90% score intervals overlap.
- 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.
Questions
Which is better, Claude Sonnet 5.5 or Qwen3.8 Max?
Claude Sonnet 5.5 has the higher public score estimate, 80.49 versus 71.69, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Claude Sonnet 5.5 or Qwen3.8 Max?
Claude Sonnet 5.5 scores higher for coding on the public lane, 79.6 to 55.6. Claude Sonnet 5.5 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, Claude Sonnet 5.5 or Qwen3.8 Max?
Claude Sonnet 5.5 leads the public agentic tasks lane, 66.1 to 64.8, with Supported evidence for both models, although the 90% intervals overlap.
Which costs less, Claude Sonnet 5.5 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, Claude Sonnet 5.5 or Qwen3.8 Max?
Both models list the same context window, 1M.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence101 rows
Agentic
Terminal-Bench 4.0
Claude Sonnet 5.570.60%SourceQwen3.8 Max—Not directly comparable
HLE w/ tools
Claude Sonnet 5.564.5%SourceQwen3.8 Max56.2%SourceClaude Sonnet 5.5 leads this result
DRACO
Claude Sonnet 5.587.0%SourceQwen3.8 Max—Not directly comparable
Terminal-Bench-Science 0.1
Claude Sonnet 5.559.9%SourceQwen3.8 Max—Not directly comparable
LAB all-pass (Harvey held-out)
Claude Sonnet 5.510.0%SourceQwen3.8 Max—Not directly comparable
LAB criterion-pass (Harvey held-out)
Claude Sonnet 5.593.1%SourceQwen3.8 Max—Not directly comparable
Toolathlon Verified Pass@3
Claude Sonnet 5.585.2%SourceQwen3.8 Max—Not directly comparable
Toolathlon Verified Pass³
Claude Sonnet 5.568.5%SourceQwen3.8 Max—Not directly comparable
Toolathlon Verified avg. turns
Claude Sonnet 5.531.6 turnsSourceQwen3.8 Max—Not directly comparable
AutomationBench (Zapier 1.0.6)
Claude Sonnet 5.544.7%SourceQwen3.8 Max—Not directly comparable
Toolathlon-Verified
Claude Sonnet 5.577.8%SourceQwen3.8 Max72.5%SourceClaude Sonnet 5.5 leads this result
Terminal-Bench 2.1
Claude Sonnet 5.5—Qwen3.8 Max86.6%SourceNot directly comparable
CoWorkBench
Claude Sonnet 5.5—Qwen3.8 Max74.8%SourceNot directly comparable
JobBench
Claude Sonnet 5.5—Qwen3.8 Max53.4%SourceNot directly comparable
skillsBench
Claude Sonnet 5.5—Qwen3.8 Max70.2%SourceNot directly comparable
Agents' Last Exam
Claude Sonnet 5.5—Qwen3.8 Max52.4%SourceNot directly comparable
AutomationBench
Claude Sonnet 5.5—Qwen3.8 Max27.3%SourceNot directly comparable
WideResearch
Claude Sonnet 5.5—Qwen3.8 Max81.9%SourceNot directly comparable
OSWorld-Verified
Claude Sonnet 5.5—Qwen3.8 Max86.1%SourceNot directly comparable
OSWorld 2.0
Claude Sonnet 5.5—Qwen3.8 Max19.4%SourceNot directly comparable
WebArena-Verified
Claude Sonnet 5.5—Qwen3.8 Max66.8%SourceNot directly comparable
AndroidWorld
Claude Sonnet 5.5—Qwen3.8 Max85.3%SourceNot directly comparable
MobileWorld
Claude Sonnet 5.5—Qwen3.8 Max77.8%SourceNot directly comparable
Terminal-Bench 2.1 (Vals)
Claude Sonnet 5.5—Qwen3.8 Max67.4%SourceNot directly comparable
Coding
FrontierCode 1.1 Main
Claude Sonnet 5.546.2%SourceQwen3.8 Max—Not directly comparable
cursorBench40
Claude Sonnet 5.555.5%SourceQwen3.8 Max—Not directly comparable
SWE-bench Pro
Claude Sonnet 5.581.3%SourceQwen3.8 Max67.7%SourceClaude Sonnet 5.5 leads this result
SWE Multilingual
Claude Sonnet 5.590.3%SourceQwen3.8 Max—Not directly comparable
SWE Multimodal
Claude Sonnet 5.554.3%SourceQwen3.8 Max—Not directly comparable
DeepSWE
Claude Sonnet 5.571.0%SourceQwen3.8 Max56.6%SourceClaude Sonnet 5.5 leads this result
FrontierCode 1.1 Extended
Claude Sonnet 5.559.1%SourceQwen3.8 Max—Not directly comparable
ProgramBench
Claude Sonnet 5.579.7%SourceQwen3.8 Max—Not directly comparable
FrontierSWE v2
Claude Sonnet 5.561.9%SourceQwen3.8 Max15.8%SourceClaude Sonnet 5.5 leads this result
Terminal-Bench 2.1
Claude Sonnet 5.5—Qwen3.8 Max86.6%SourceNot directly comparable
NL2Repo
Claude Sonnet 5.5—Qwen3.8 Max55.9%SourceNot directly comparable
FrontierSWE
Claude Sonnet 5.5—Qwen3.8 Max73.5%SourceNot directly comparable
MLS-Bench Lite
Claude Sonnet 5.5—Qwen3.8 Max41.0%SourceNot directly comparable
PaperBench
Claude Sonnet 5.5—Qwen3.8 Max93.0%SourceNot directly comparable
VulcanBench v3
Claude Sonnet 5.5—Qwen3.8 Max81.2%SourceNot directly comparable
OpenHarmony Bench
Claude Sonnet 5.5—Qwen3.8 Max60.8%SourceNot directly comparable
LiveCodeBench (Vals)
Claude Sonnet 5.5—Qwen3.8 Max87.9%SourceNot directly comparable
SWE-bench (Vals)
Claude Sonnet 5.5—Qwen3.8 Max85.6%SourceNot directly comparable
Reasoning
Multimodal
Chartography (no tools)
Claude Sonnet 5.561.6%SourceQwen3.8 Max—Not directly comparable
Chartography (tools)
Claude Sonnet 5.590.2%SourceQwen3.8 Max—Not directly comparable
BenchCAD Vision2Code (no tools)
Claude Sonnet 5.50.747SourceQwen3.8 Max—Not directly comparable
BenchCAD Vision2Code (tools)
Claude Sonnet 5.50.963SourceQwen3.8 Max—Not directly comparable
Biomedical image analysis
Claude Sonnet 5.572.2%SourceQwen3.8 Max—Not directly comparable
OfficeQA
Claude Sonnet 5.576.9%SourceQwen3.8 Max—Not directly comparable
OfficeQA Pro
Claude Sonnet 5.565.6%SourceQwen3.8 Max—Not directly comparable
MMMU-Pro
Claude Sonnet 5.5—Qwen3.8 Max82.3%SourceNot directly comparable
MathVision
Claude Sonnet 5.5—Qwen3.8 Max95.2%SourceNot directly comparable
MathVision w/ Python
Claude Sonnet 5.5—Qwen3.8 Max97.7%SourceNot directly comparable
BabyVision
Claude Sonnet 5.5—Qwen3.8 Max82.0%SourceNot directly comparable
BabyVision w/ Python
Claude Sonnet 5.5—Qwen3.8 Max91.3%SourceNot directly comparable
ZeroBench
Claude Sonnet 5.5—Qwen3.8 Max24.0%SourceNot directly comparable
ZeroBench w/ Python
Claude Sonnet 5.5—Qwen3.8 Max49.0%SourceNot directly comparable
MedXpertQA (MM)
Claude Sonnet 5.5—Qwen3.8 Max80.4%SourceNot directly comparable
ScreenSpot Pro
Claude Sonnet 5.5—Qwen3.8 Max84.5%SourceNot directly comparable
Vision2Web
Claude Sonnet 5.5—Qwen3.8 Max69.0%SourceNot directly comparable
CharXiv w/o tools
Claude Sonnet 5.5—Qwen3.8 Max88.4%SourceNot directly comparable
CharXiv
Claude Sonnet 5.5—Qwen3.8 Max93.5%SourceNot directly comparable
OmniDocBench 1.5
Claude Sonnet 5.5—Qwen3.8 Max92.1%SourceNot directly comparable
OCRBench V2
Claude Sonnet 5.5—Qwen3.8 Max74.2%SourceNot directly comparable
CC-OCR
Claude Sonnet 5.5—Qwen3.8 Max79.6%SourceNot directly comparable
RealWorldQA
Claude Sonnet 5.5—Qwen3.8 Max88.0%SourceNot directly comparable
ERQA
Claude Sonnet 5.5—Qwen3.8 Max77.8%SourceNot directly comparable
SimpleVQA
Claude Sonnet 5.5—Qwen3.8 Max75.0%SourceNot directly comparable
PerceptionBench
Claude Sonnet 5.5—Qwen3.8 Max63.5%SourceNot directly comparable
Video-MME (with subtitle)
Claude Sonnet 5.5—Qwen3.8 Max90.4%SourceNot directly comparable
VideoMMMU
Claude Sonnet 5.5—Qwen3.8 Max88.7%SourceNot directly comparable
MMVU
Claude Sonnet 5.5—Qwen3.8 Max82.4%SourceNot directly comparable
MLVU (M-Avg)
Claude Sonnet 5.5—Qwen3.8 Max90.8%SourceNot directly comparable
LVBench
Claude Sonnet 5.5—Qwen3.8 Max81.8%SourceNot directly comparable
Knowledge
HealthBench (raw)
Claude Sonnet 5.569.4%SourceQwen3.8 Max—Not directly comparable
HealthBench (length-adjusted)
Claude Sonnet 5.565.4%SourceQwen3.8 Max—Not directly comparable
HealthBench Professional (raw)
Claude Sonnet 5.577.1%SourceQwen3.8 Max—Not directly comparable
HealthBench Professional
Claude Sonnet 5.569.2%SourceQwen3.8 Max—Not directly comparable
BioMysteryBench (human-solvable)
Claude Sonnet 5.589.2%SourceQwen3.8 Max—Not directly comparable
BioMysteryBench (human-difficult)
Claude Sonnet 5.544.7%SourceQwen3.8 Max—Not directly comparable
SpatialBench Verified
Claude Sonnet 5.572.5%SourceQwen3.8 Max—Not directly comparable
SingleCellBench
Claude Sonnet 5.559.1%SourceQwen3.8 Max—Not directly comparable
Protein Design
Claude Sonnet 5.551.0%SourceQwen3.8 Max—Not directly comparable
Morphology-to-molecule matching
Claude Sonnet 5.525.0%SourceQwen3.8 Max—Not directly comparable
Medicinal chemistry
Claude Sonnet 5.565.3%SourceQwen3.8 Max—Not directly comparable
Protein Design library ranking
Claude Sonnet 5.554.8%SourceQwen3.8 Max—Not directly comparable
De novo protein-binder design
Claude Sonnet 5.582.3%SourceQwen3.8 Max—Not directly comparable
Protocols (troubleshooting)
Claude Sonnet 5.567.3%SourceQwen3.8 Max—Not directly comparable
Protocols (understanding)
Claude Sonnet 5.566.6%SourceQwen3.8 Max—Not directly comparable
HLE w/o tools
Claude Sonnet 5.556.9%SourceQwen3.8 Max43.6%SourceClaude Sonnet 5.5 leads this result
GPQA
Claude Sonnet 5.5—Qwen3.8 Max92.6%SourceNot directly comparable
GPQA-D
Claude Sonnet 5.5—Qwen3.8 Max92.6%SourceNot directly comparable
HLE
Claude Sonnet 5.5—Qwen3.8 Max43.6%SourceNot directly comparable
GPQA Diamond (Vals)
Claude Sonnet 5.5—Qwen3.8 Max93.7%SourceNot directly comparable
MMLU-Pro (Vals)
Claude Sonnet 5.5—Qwen3.8 Max88.6%SourceNot directly comparable
Multilingual
Instruction following
IFBench
Claude Sonnet 5.5—Qwen3.8 Max82.8%SourceNot directly comparable
Math
101 public results · 6 shared
Watch Claude Sonnet 5.5 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.
Last updated September 28, 2026