Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

See the free Radar Brief
BenchLM
Anthropic logo
Model A
Claude 4.1 Opus

Anthropic

44.39/100

Supported · Public rank #169

90% interval 41.5–47.3

Claude 4.1 Opus vs Claude Opus 5

Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Anthropic logo
Model B
Claude Opus 5

Anthropic

80.66/100

Supported · Public rank #4

90% interval 78.3–83.0

Decision reading

Claude Opus 5 has the higher public score, 80.66 versus 44.39, and the 90% score intervals do not overlap.

1 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Share or export

Share on XLinkedInSocial cardCSVJSON

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

    Claude Opus 5

    Claude Opus 5 has the larger documented context window.

    Confidence: documented

  • Chat turn cost

    1K fresh input + 500 output tokens

    Claude Opus 5

    Claude Opus 5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

Show secondary and unsupported calls

  • Repository review cost

    50K fresh input + 3K output tokens

    Claude Opus 5

    Claude Opus 5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

  • Coding work

    Code generation, repair, and software-engineering tasks

    Not enough matched evidence

    Claude 4.1 Opus 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

    Claude 4.1 Opus is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.

    Confidence: limited

  • 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. Claude 4.1 Opus does not fit this workload in one request. Claude 4.1 Opus has no published cached-input rate, so cached tokens use its listed input rate.

    Confidence: rate-fallback

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

Shared results
1
Claude 4.1 Opus only
1
Claude Opus 5 only
72
Like-for-like categories
0 / 8

2 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
Claude 4.1 Opus
45.1
Estimated · #95/151
Claude Opus 5
77.4
Supported · #2/151
Basis
BenchAlign lane · 1 vs 19 public rows
Reading
Directional only

Coding

Directional only
Claude 4.1 Opus
45.1
Estimated · #111/183
Claude Opus 5
75.6
Supported · #3/183
Basis
BenchAlign lane · 1 vs 16 public rows
Reading
Directional only

Reasoning

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
75.6
#13/22
Basis
Provisional lane · 0 vs 2 weighted rows
Reading
Not comparable

Knowledge

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
82.1
Supported · #3/181
Basis
BenchAlign lane · 0 vs 19 public rows
Reading
Not comparable

Math

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
Not ranked
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

Multilingual

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
Not ranked
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

Multimodal

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
88.7
#2/48
Basis
Provisional lane · 0 vs 1 weighted rows
Reading
Not comparable

Instruction following

Not comparable
Claude 4.1 Opus
Not ranked
Claude Opus 5
Not ranked
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
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.

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 4.1 Opus
$0.0525
Fits in one request
Claude Opus 5
$0.0175
Fits in one request

Claude Opus 5 has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Claude 4.1 Opus
$0.975
Fits in one request
Claude Opus 5
$0.325
Fits in one request

Claude Opus 5 has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Claude 4.1 Opus
$4.05
Does not fit in one request
Cached input priced at the published list-input rate
Claude Opus 5
$0.45
Fits in one request

Claude 4.1 Opus does not fit this workload in one request. Claude 4.1 Opus has no published cached-input rate, so cached tokens use its listed input rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

Context window

Maximum documented context; output-token limits may be lower.

Claude 4.1 Opus

200K

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Claude 4.1 Opus

Not published

Claude Opus 5

$0.5 per 1M cached input tokens

Claude API pricing

Reasoning profile

Claude 4.1 Opus

Non-Reasoning

Claude Opus 5

Reasoning

Weight access

Claude 4.1 Opus

Proprietary

Claude Opus 5

Proprietary

License

Claude 4.1 Opus

Proprietary

Claude Opus 5

Proprietary

Release date

Claude 4.1 Opus

2025-08-01

Claude Opus 5

2026-07-24

If you already use one of these models
Deployment change
Both entries list Anthropic as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
Quality signal
Claude Opus 5 has the higher public score, 80.66 versus 44.39, and the 90% score intervals do not overlap.
Workload cost
Repository review: $0.975 vs $0.325. Cache-heavy agent loop: $4.05 vs $0.45.
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 evidence74 rows

Agentic

  • JobBench

    Claude 4.1 Opus21.9%
    Source
    Claude Opus 5—

    Not directly comparable

  • Terminal-Bench 3.0

    Claude 4.1 Opus—
    Claude Opus 542.7%
    Source

    Not directly comparable

  • BrowseComp

    Claude 4.1 Opus—
    Claude Opus 590.8%
    Source

    Not directly comparable

  • HLE w/ tools

    Claude 4.1 Opus—
    Claude Opus 564.7%
    Source

    Not directly comparable

  • DeepSearchQA

    Claude 4.1 Opus—
    Claude Opus 595.0%
    Source

    Not directly comparable

  • DRACO

    Claude 4.1 Opus—
    Claude Opus 588.6%
    Source

    Not directly comparable

  • BrowseComp (10-agent, prerelease)

    Claude 4.1 Opus—
    Claude Opus 593.6%
    Source

    Not directly comparable

  • OSWorld 2.0

    Claude 4.1 Opus—
    Claude Opus 570.6%
    Source

    Not directly comparable

  • MCP Atlas

    Claude 4.1 Opus—
    Claude Opus 585.8%
    Source

    Not directly comparable

  • MCP-Atlas claim coverage

    Claude 4.1 Opus—
    Claude Opus 589.1%
    Source

    Not directly comparable

  • LAB all-pass (Anthropic harness)

    Claude 4.1 Opus—
    Claude Opus 523.58%
    Source

    Not directly comparable

  • LAB criterion-pass (Anthropic harness)

    Claude 4.1 Opus—
    Claude Opus 593.74%
    Source

    Not directly comparable

  • LAB all-pass (Harvey held-out)

    Claude 4.1 Opus—
    Claude Opus 511.7%
    Source

    Not directly comparable

  • LAB criterion-pass (Harvey held-out)

    Claude 4.1 Opus—
    Claude Opus 594.1%
    Source

    Not directly comparable

  • Toolathlon-Verified

    Claude 4.1 Opus—
    Claude Opus 580.6%
    Source

    Not directly comparable

  • Toolathlon Verified Pass@3

    Claude 4.1 Opus—
    Claude Opus 587.0%
    Source

    Not directly comparable

  • Toolathlon Verified Pass³

    Claude 4.1 Opus—
    Claude Opus 573.1%
    Source

    Not directly comparable

  • Toolathlon Verified avg. turns

    Claude 4.1 Opus—
    Claude Opus 523.5 turns
    Source

    Not directly comparable

  • AutomationBench

    Claude 4.1 Opus—
    Claude Opus 526.0%
    Source

    Not directly comparable

  • Terminal-Bench 2.1 (Vals)

    Claude 4.1 Opus—
    Claude Opus 584.6%
    Source

    Not directly comparable

Coding

  • SWE-bench Verified

    Claude 4.1 Opus74.5%
    Source
    Claude Opus 596%
    Source

    Claude Opus 5 leads this result

  • Bug Hunt Bench

    Claude 4.1 Opus—
    Claude Opus 527 fixes
    Source

    Not directly comparable

  • SWE-bench Pro

    Claude 4.1 Opus—
    Claude Opus 579.2%
    Source

    Not directly comparable

  • SWE Multilingual

    Claude 4.1 Opus—
    Claude Opus 589.5%
    Source

    Not directly comparable

  • SWE Multimodal

    Claude 4.1 Opus—
    Claude Opus 559.4%
    Source

    Not directly comparable

  • deepSwe

    Claude 4.1 Opus—
    Claude Opus 568.8%
    Source

    Not directly comparable

  • FrontierCode 1.1 Main

    Claude 4.1 Opus—
    Claude Opus 553.4%
    Source

    Not directly comparable

  • FrontierCode 1.1 Extended

    Claude 4.1 Opus—
    Claude Opus 563.6%
    Source

    Not directly comparable

  • FrontierSWE v2

    Claude 4.1 Opus—
    Claude Opus 552.0%
    Source

    Not directly comparable

  • ProgramBench (episode 1)

    Claude 4.1 Opus—
    Claude Opus 583.0%
    Source

    Not directly comparable

  • ProgramBench

    Claude 4.1 Opus—
    Claude Opus 593.0%
    Source

    Not directly comparable

  • cursorBench32

    Claude 4.1 Opus—
    Claude Opus 570.0%
    Source

    Not directly comparable

  • VulcanBench v3

    Claude 4.1 Opus—
    Claude Opus 587.0%
    Source

    Not directly comparable

  • VulcanBench CII v1

    Claude 4.1 Opus—
    Claude Opus 596.4%
    Source

    Not directly comparable

  • LiveCodeBench (Vals)

    Claude 4.1 Opus—
    Claude Opus 589.0%
    Source

    Not directly comparable

  • SWE-bench (Vals)

    Claude 4.1 Opus—
    Claude Opus 597.0%
    Source

    Not directly comparable

Reasoning

  • ARC-AGI-1

    Claude 4.1 Opus—
    Claude Opus 597.50%
    Source

    Not directly comparable

  • ARC-AGI-2

    Claude 4.1 Opus—
    Claude Opus 590.4%
    Source

    Not directly comparable

  • ARC-AGI-3

    Claude 4.1 Opus—
    Claude Opus 530.2%
    Source

    Not directly comparable

Knowledge

  • HLE

    Claude 4.1 Opus—
    Claude Opus 564.7%
    Source

    Not directly comparable

  • HLE w/o tools

    Claude 4.1 Opus—
    Claude Opus 556.3%
    Source

    Not directly comparable

  • HLE-Verified

    Claude 4.1 Opus—
    Claude Opus 554.4%
    Source

    Not directly comparable

  • LABBench2

    Claude 4.1 Opus—
    Claude Opus 584.2%
    Source

    Not directly comparable

  • HealthBench (raw)

    Claude 4.1 Opus—
    Claude Opus 567.1%
    Source

    Not directly comparable

  • HealthBench (length-adjusted)

    Claude 4.1 Opus—
    Claude Opus 557.8%
    Source

    Not directly comparable

  • HealthBench Professional

    Claude 4.1 Opus—
    Claude Opus 559.8%
    Source

    Not directly comparable

  • HealthBench Professional (raw)

    Claude 4.1 Opus—
    Claude Opus 573.4%
    Source

    Not directly comparable

  • BioMysteryBench (human-solvable)

    Claude 4.1 Opus—
    Claude Opus 590.1%
    Source

    Not directly comparable

  • BioMysteryBench (human-difficult)

    Claude 4.1 Opus—
    Claude Opus 549.4%
    Source

    Not directly comparable

  • SpatialBench Verified

    Claude 4.1 Opus—
    Claude Opus 572.5%
    Source

    Not directly comparable

  • SingleCellBench

    Claude 4.1 Opus—
    Claude Opus 560.6%
    Source

    Not directly comparable

  • ProteinGym Hard

    Claude 4.1 Opus—
    Claude Opus 547.7%
    Source

    Not directly comparable

  • Protein Design

    Claude 4.1 Opus—
    Claude Opus 542.5%
    Source

    Not directly comparable

  • Organic chemistry V2

    Claude 4.1 Opus—
    Claude Opus 561.6%
    Source

    Not directly comparable

  • Protocols (troubleshooting)

    Claude 4.1 Opus—
    Claude Opus 561.1%
    Source

    Not directly comparable

  • Protocols (understanding)

    Claude 4.1 Opus—
    Claude Opus 578.4%
    Source

    Not directly comparable

  • GPQA Diamond (Vals)

    Claude 4.1 Opus—
    Claude Opus 593.4%
    Source

    Not directly comparable

  • MMLU-Pro (Vals)

    Claude 4.1 Opus—
    Claude Opus 591.6%
    Source

    Not directly comparable

Math

  • IMO 2026

    Claude 4.1 Opus—
    Claude Opus 542/42
    Source

    Not directly comparable

  • RiemannBench (no tools)

    Claude 4.1 Opus—
    Claude Opus 560.0%
    Source

    Not directly comparable

  • RiemannBench (tools)

    Claude 4.1 Opus—
    Claude Opus 579.0%
    Source

    Not directly comparable

  • ArXivMath Jun. 2026 (no tools)

    Claude 4.1 Opus—
    Claude Opus 590.8%
    Source

    Not directly comparable

  • ArXivMath Jun. 2026 (tools)

    Claude 4.1 Opus—
    Claude Opus 591.3%
    Source

    Not directly comparable

Multilingual

  • GMMLU

    Claude 4.1 Opus—
    Claude Opus 592.5%
    Source

    Not directly comparable

  • MILU

    Claude 4.1 Opus—
    Claude Opus 592.1%
    Source

    Not directly comparable

  • INCLUDE

    Claude 4.1 Opus—
    Claude Opus 589.8%
    Source

    Not directly comparable

Multimodal

  • Chartography (no tools)

    Claude 4.1 Opus—
    Claude Opus 529.6%
    Source

    Not directly comparable

  • Chartography (tools)

    Claude 4.1 Opus—
    Claude Opus 583.0%
    Source

    Not directly comparable

  • BenchCAD Vision2Code (no tools)

    Claude 4.1 Opus—
    Claude Opus 50.366
    Source

    Not directly comparable

  • BenchCAD Vision2Code (tools)

    Claude 4.1 Opus—
    Claude Opus 50.821
    Source

    Not directly comparable

  • GDP.pdf (no tools)

    Claude 4.1 Opus—
    Claude Opus 583.4%
    Source

    Not directly comparable

  • GDP.pdf (tools)

    Claude 4.1 Opus—
    Claude Opus 585.5%
    Source

    Not directly comparable

  • OfficeQA

    Claude 4.1 Opus—
    Claude Opus 578.1%
    Source

    Not directly comparable

  • OfficeQA Pro

    Claude 4.1 Opus—
    Claude Opus 566.9%
    Source

    Not directly comparable

Frequently asked questions

Which is better, Claude 4.1 Opus or Claude Opus 5?

Claude Opus 5 has the higher public score, 80.66 versus 44.39, 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, Claude 4.1 Opus or Claude Opus 5?

Claude Opus 5 scores higher for coding on the public lane, 75.6 to 45.1. Claude 4.1 Opus 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 4.1 Opus or Claude Opus 5?

Claude Opus 5 scores higher for agentic tasks on the public lane, 77.4 to 45.1. Claude 4.1 Opus 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, Claude 4.1 Opus or Claude Opus 5?

For the stated presets, chat costs $0.0525 on Claude 4.1 Opus and $0.0175 on Claude Opus 5; repository review costs $0.975 and $0.325; the cache-heavy agent loop costs $4.05 and $0.45. Claude 4.1 Opus does not fit this workload in one request. Claude 4.1 Opus has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Claude 4.1 Opus or Claude Opus 5?

Claude Opus 5 has the larger documented context window: 1M, compared with 200K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated September 4, 2026

Watch Claude 4.1 Opus vs Claude Opus 5

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.