Agentic
Like-for-like- Claude Opus 5
- 90.8
- dots3-note Preview
- 83.3
- Weighted basis
- 1 vs 1 rows
- Reading
- Claude Opus 5 leads
Claude Opus 5 vs dots3-note Preview
Updated August 14, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
10 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.
Coding work
Code generation, repair, and software-engineering tasks
Claude Opus 5
Claude Opus 5 leads on the same 2 weighted benchmark rows.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Claude Opus 5
Claude Opus 5 leads on the same 1 weighted benchmark row.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Claude Opus 5
Claude Opus 5 has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
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.
53 Claude Opus 5 only
10 Shared
21 dots3-note Preview only
- Shared results
- 10
- Claude Opus 5 only
- 53
- dots3-note Preview only
- 21
- Like-for-like categories
- 4 / 8
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.
Coding
Like-for-like- Claude Opus 5
- 89.5
- dots3-note Preview
- 71.7
- Weighted basis
- 2 vs 2 rows
- Reading
- Claude Opus 5 leads
Reasoning
Like-for-like- Claude Opus 5
- 90.4
- dots3-note Preview
- 81.4
- Weighted basis
- 1 vs 1 rows
- Reading
- Claude Opus 5 leads
Knowledge
Like-for-like- Claude Opus 5
- 64.7
- dots3-note Preview
- 52.6
- Weighted basis
- 1 vs 1 rows
- Reading
- Claude Opus 5 leads
Math
Not comparable- Claude Opus 5
- Not measured
- dots3-note Preview
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Opus 5
- Not measured
- dots3-note Preview
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Opus 5
- 66.9
- dots3-note Preview
- 79.1
- Weighted basis
- 1 vs 1 rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Opus 5
- Not measured
- dots3-note Preview
- 85.1
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
| Category | Claude Opus 5 | dots3-note Preview | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | 90.8 | 83.3 | Like-for-like1 vs 1 rows | Claude Opus 5 leads |
| Coding | 89.5 | 71.7 | Like-for-like2 vs 2 rows | Claude Opus 5 leads |
| Reasoning | 90.4 | 81.4 | Like-for-like1 vs 1 rows | Claude Opus 5 leads |
| Knowledge | 64.7 | 52.6 | Like-for-like1 vs 1 rows | Claude Opus 5 leads |
| 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 | 66.9 | 79.1 | Not comparable1 vs 1 rows | Not comparable |
| Instruction following | Not measured | 85.1 | Not comparable0 vs 2 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.
SWE-bench Pro
Coding
Claude Opus 5: 79.2%dots3-note Preview: 61%Normalized gap 18.2Claude Opus 5 source dots3-note Preview sourceSWE-bench Verified
Coding
Claude Opus 5: 96%dots3-note Preview: 78.4%Normalized gap 17.6Claude Opus 5 source dots3-note Preview sourceHLE
Knowledge
Claude Opus 5: 64.7%dots3-note Preview: 52.6%Normalized gap 12.1Claude Opus 5 source dots3-note Preview sourceARC-AGI-2
Reasoning
Claude Opus 5: 90.4%dots3-note Preview: 81.4%Normalized gap 9.0Claude Opus 5 source dots3-note Preview sourceBrowseComp
Agentic
Claude Opus 5: 90.8%dots3-note Preview: 83.3%Normalized gap 7.5Claude Opus 5 source dots3-note Preview 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
- Claude Opus 5
- $0.0175
- Fits in one request
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
dots3-note Preview has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Claude Opus 5
- $0.325
- Fits in one request
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
dots3-note Preview has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Claude Opus 5
- $0.45
- Fits in one request
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
dots3-note Preview 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.
SpecificationClaude Opus 5dots3-note Preview
Context window
Maximum documented context; output-token limits may be lower.
Claude Opus 5
dots3-note Preview
API model ID
Claude Opus 5
claude-opus-5
Anthropic model overviewdots3-note Preview
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Opus 5
$0.5 per 1M cached input tokens
Claude API pricingdots3-note Preview
No comparable hosted API rate
dots3-note Preview model cardDocumented inputs
Claude Opus 5
text, image
Anthropic model overviewdots3-note Preview
Not sourced
Documented outputs
Claude Opus 5
dots3-note Preview
Not sourced
Provider availability
Claude Opus 5
Generally Available · Claude API
Anthropic model overviewdots3-note Preview
Not sourced
Reasoning profile
Claude Opus 5
Reasoning
dots3-note Preview
Reasoning
Weight access
Claude Opus 5
Proprietary
dots3-note Preview
Open Weight
License
Claude Opus 5
Proprietary
dots3-note Preview
Open Weight
Release date
Claude Opus 5
2026-07-24
dots3-note Preview
2026-08-14
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
- Claude Opus 5 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 evidence84 rows
Agentic
Terminal-Bench 3.0
Claude Opus 542.7%Sourcedots3-note Preview—Not directly comparable
BrowseComp
Claude Opus 590.8%Sourcedots3-note Preview83.3%SourceClaude Opus 5 leads this result
HLE w/ tools
Claude Opus 564.7%Sourcedots3-note Preview52.6%SourceClaude Opus 5 leads this result
DeepSearchQA
Claude Opus 595.0%Sourcedots3-note Preview92.1%SourceClaude Opus 5 leads this result
DRACO
Claude Opus 588.6%Sourcedots3-note Preview—Not directly comparable
BrowseComp (10-agent, prerelease)
Claude Opus 593.6%Sourcedots3-note Preview—Not directly comparable
OSWorld 2.0
Claude Opus 570.6%Sourcedots3-note Preview—Not directly comparable
MCP Atlas
Claude Opus 585.8%Sourcedots3-note Preview—Not directly comparable
MCP-Atlas claim coverage
Claude Opus 589.1%Sourcedots3-note Preview—Not directly comparable
LAB all-pass (Anthropic harness)
Claude Opus 523.58%Sourcedots3-note Preview—Not directly comparable
LAB criterion-pass (Anthropic harness)
Claude Opus 593.74%Sourcedots3-note Preview—Not directly comparable
LAB all-pass (Harvey held-out)
Claude Opus 511.7%Sourcedots3-note Preview—Not directly comparable
LAB criterion-pass (Harvey held-out)
Claude Opus 594.1%Sourcedots3-note Preview—Not directly comparable
Toolathlon-Verified
Claude Opus 580.6%Sourcedots3-note Preview55.6%SourceClaude Opus 5 leads this result
Toolathlon Verified Pass@3
Claude Opus 587.0%Sourcedots3-note Preview—Not directly comparable
Toolathlon Verified Pass³
Claude Opus 573.1%Sourcedots3-note Preview—Not directly comparable
Toolathlon Verified avg. turns
Claude Opus 523.5 turnsSourcedots3-note Preview—Not directly comparable
AutomationBench
Claude Opus 526.0%Sourcedots3-note Preview—Not directly comparable
Claw-Eval
Claude Opus 5—dots3-note Preview73.4%SourceNot directly comparable
Terminal-Bench 2.1
Claude Opus 5—dots3-note Preview75.1%SourceNot directly comparable
skillsBench
Claude Opus 5—dots3-note Preview52.8%SourceNot directly comparable
APEX-Agents
Claude Opus 5—dots3-note Preview30.8%SourceNot directly comparable
WideResearch
Claude Opus 5—dots3-note Preview78.9%SourceNot directly comparable
Coding
SWE-bench Verified
Claude Opus 596%Sourcedots3-note Preview78.4%SourceClaude Opus 5 leads this result
SWE-bench Pro
Claude Opus 579.2%Sourcedots3-note Preview61%SourceClaude Opus 5 leads this result
SWE Multilingual
Claude Opus 589.5%Sourcedots3-note Preview75.7%SourceClaude Opus 5 leads this result
SWE Multimodal
Claude Opus 559.4%Sourcedots3-note Preview—Not directly comparable
deepSwe
Claude Opus 568.8%Sourcedots3-note Preview—Not directly comparable
FrontierCode 1.1 Main
Claude Opus 553.4%Sourcedots3-note Preview—Not directly comparable
FrontierCode 1.1 Extended
Claude Opus 563.6%Sourcedots3-note Preview—Not directly comparable
ProgramBench (episode 1)
Claude Opus 583.0%Sourcedots3-note Preview—Not directly comparable
ProgramBench
Claude Opus 593.0%Sourcedots3-note Preview—Not directly comparable
cursorBench32
Claude Opus 570.0%Sourcedots3-note Preview—Not directly comparable
VulcanBench v3
Claude Opus 587.0%Sourcedots3-note Preview—Not directly comparable
Codeforces
Claude Opus 5—dots3-note Preview3056.0SourceNot directly comparable
LiveCodeBench v6
Claude Opus 5—dots3-note Preview91.5%SourceNot directly comparable
Terminal-Bench 2.1
Claude Opus 5—dots3-note Preview75.1%SourceNot directly comparable
NL2Repo
Claude Opus 5—dots3-note Preview49.8%SourceNot directly comparable
Reasoning
Knowledge
HLE
Claude Opus 564.7%Sourcedots3-note Preview52.6%SourceClaude Opus 5 leads this result
HLE w/o tools
Claude Opus 556.3%Sourcedots3-note Preview—Not directly comparable
HealthBench (raw)
Claude Opus 567.1%Sourcedots3-note Preview—Not directly comparable
HealthBench (length-adjusted)
Claude Opus 557.8%Sourcedots3-note Preview—Not directly comparable
HealthBench Professional
Claude Opus 559.8%Sourcedots3-note Preview—Not directly comparable
HealthBench Professional (raw)
Claude Opus 573.4%Sourcedots3-note Preview—Not directly comparable
BioMysteryBench (human-solvable)
Claude Opus 590.1%Sourcedots3-note Preview—Not directly comparable
BioMysteryBench (human-difficult)
Claude Opus 549.4%Sourcedots3-note Preview—Not directly comparable
SpatialBench Verified
Claude Opus 572.5%Sourcedots3-note Preview—Not directly comparable
SingleCellBench
Claude Opus 560.6%Sourcedots3-note Preview—Not directly comparable
ProteinGym Hard
Claude Opus 547.7%Sourcedots3-note Preview—Not directly comparable
Protein Design
Claude Opus 542.5%Sourcedots3-note Preview—Not directly comparable
Organic chemistry V2
Claude Opus 561.6%Sourcedots3-note Preview—Not directly comparable
Protocols (troubleshooting)
Claude Opus 561.1%Sourcedots3-note Preview—Not directly comparable
Protocols (understanding)
Claude Opus 578.4%Sourcedots3-note Preview—Not directly comparable
Math
IMO 2026
Claude Opus 542/42Sourcedots3-note Preview—Not directly comparable
RiemannBench (no tools)
Claude Opus 560.0%Sourcedots3-note Preview—Not directly comparable
RiemannBench (tools)
Claude Opus 579.0%Sourcedots3-note Preview—Not directly comparable
ArXivMath Jun. 2026 (no tools)
Claude Opus 590.8%Sourcedots3-note Preview—Not directly comparable
ArXivMath Jun. 2026 (tools)
Claude Opus 591.3%Sourcedots3-note Preview—Not directly comparable
IMOAnswerBench
Claude Opus 5—dots3-note Preview90.9%SourceNot directly comparable
Multilingual
Multimodal
Chartography (no tools)
Claude Opus 529.6%Sourcedots3-note Preview—Not directly comparable
Chartography (tools)
Claude Opus 583.0%Sourcedots3-note Preview—Not directly comparable
BenchCAD Vision2Code (no tools)
Claude Opus 50.366Sourcedots3-note Preview—Not directly comparable
BenchCAD Vision2Code (tools)
Claude Opus 50.821Sourcedots3-note Preview—Not directly comparable
GDP.pdf (no tools)
Claude Opus 583.4%Sourcedots3-note Preview60.7%SourceClaude Opus 5 leads this result
GDP.pdf (tools)
Claude Opus 585.5%Sourcedots3-note Preview—Not directly comparable
OfficeQA
Claude Opus 578.1%Sourcedots3-note Preview—Not directly comparable
OfficeQA Pro
Claude Opus 566.9%Sourcedots3-note Preview—Not directly comparable
SimpleVQA
Claude Opus 5—dots3-note Preview72.5%SourceNot directly comparable
MMMU-Pro
Claude Opus 5—dots3-note Preview79.1%SourceNot directly comparable
MathVision
Claude Opus 5—dots3-note Preview87.7%SourceNot directly comparable
ZeroBench
Claude Opus 5—dots3-note Preview19.0%SourceNot directly comparable
CharXiv w/o tools
Claude Opus 5—dots3-note Preview83.1%SourceNot directly comparable
PerceptionBench
Claude Opus 5—dots3-note Preview53.4%SourceNot directly comparable
BabyVision
Claude Opus 5—dots3-note Preview50.0%SourceNot directly comparable
MMVU
Claude Opus 5—dots3-note Preview79.9%SourceNot directly comparable
VideoMMMU
Claude Opus 5—dots3-note Preview86.8%SourceNot directly comparable
Instruction following
Frequently asked questions
Which is better, Claude Opus 5 or dots3-note Preview?
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, Claude Opus 5 or dots3-note Preview?
Claude Opus 5 leads the like-for-like coding comparison across 2 shared weighted benchmark rows.
Which is better for agentic tasks, Claude Opus 5 or dots3-note Preview?
Claude Opus 5 leads the like-for-like agentic tasks comparison across 1 shared weighted benchmark row.
Which costs less, Claude Opus 5 or dots3-note Preview?
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 Opus 5 or dots3-note Preview?
Claude Opus 5 has the larger documented context window: 1M, compared with 512K.
Related comparisons
Last updated August 14, 2026
Watch Claude Opus 5 vs dots3-note Preview
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.