Skip to content
AI Atlas
ModelActiveClosed / proprietary

Claude Opus 5

Anthropicfamily · Claudeplatform.claude.com/docs/en/models/opus-5/overvi

For complex agentic coding and enterprise work

Open in Graph
data quality77

Updated 2 h ago · first seen 11 Sept 2026

model_01M293V7EQDY9HVA5KVN8PB80A

Overview

Identity

Canonical model
Yesidentity confidence: highOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
Official checkpoints
None recorded — closed weightsofficial_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
Artifacts
None recordedSeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
Provider deployments
1
API aliases
anthropic/claude-opus-5claude-opus-5claude-opus-5-highclaude-opus-5-lowclaude-opus-5-mediumclaude-opus-5-xhighIdentifiers under which providers and evaluators refer to this model.
Folded evaluation variants
5Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.

Openness

Closed / proprietary— weights not available; 7 dimensions unknown.

Weights are not available; the model is reachable only through an API or a product.

  • Weights

    No

  • Inference code

    —

  • Training code

    —

  • Training data

    —

  • Dataset

    —

  • Commercial use

    —

  • Redistribution

    —

  • Derivatives

    —

dimensions marked null are unknown, not false

Key facts

Release date

Source:Claude Developer Platform docsT1observed 12 h agohigh

Status

Source:Claude Developer Platform docsT1observed 12 h agohigh

Version

Source:Claude Developer Platform docsT2observed 12 h agomediumLLM-extracted

Knowledge cutoff

Source:Claude Developer Platform docsT1observed 12 h agohigh

Retirement date

Source:Claude Developer Platform docsT1observed 12 h agohigh

Official page
platform.claude.com/docs/en/models/opus-5/overview

Source:Claude Developer Platform docsT1observed 12 h agohigh

API model id

Source:Claude Developer Platform docsT1observed 12 h agohigh

Openrouter id

Source:OpenRouter public model & pricing listingT2observed 12 h agomedium

Architecture

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 12 h agomedium

Training data cutoff

Source:Claude Developer Platform docsT1observed 12 h agohigh

Capabilities

Modalities

Modalities
textimage
Input
textimage
Output
text

Capabilities

  • Tool calling

    Yes

    Claude Developer Platform docs · T1

  • Structured output

    Yes

    OpenRouter public model & pricing listing · T2

  • Reasoning

    Yes

    OpenRouter public model & pricing listing · T2

  • Vision

    Yes

    Claude Developer Platform docs · T1

  • Audio

    Unavailable

  • Fine-tuning available

    Unavailable

Context window

Source:Claude Developer Platform docsT1observed 12 h agohigh

Max output

Source:Claude Developer Platform docsT1observed 12 h agohigh

Knowledge cutoff

Source:Claude Developer Platform docsT1observed 12 h agohigh

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 12 h agomedium

Benchmarks43

Compare with another model →

Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric

Benchmark results grouped by comparability group
Benchmark · groupBest scoreTrustConfigurationResultsvs leaderEvaluatedSource
Terminal-Benchagentic · accuracy · variant=v2.1 · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisvariantv2.15non-primary groupobs. 11 Sept 2026artificialanalysis.aiT2
Terminal-Benchagentic · accuracy · variant=v4.0 · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisvariantv4.05non-primary groupobs. 11 Sept 2026artificialanalysis.aiT2
LiveBench Agentic Codingcoding · average score · variant=Agentic CodingOfficial boardvariantAgentic Codingreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−12.1 ptvs DeepSeek-V4.1-Flash25 Jun 2026livebench.aiT2
LiveBench Codingcoding · average score · variant=CodingOfficial boardvariantCodingreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−4.93 ptvs Claude Fable 5.125 Jun 2026livebench.aiT2
SciCodecoding · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysis5−6.71 ptvs Claude Fable 5.1obs. 11 Sept 2026artificialanalysis.aiT2
Artificial Analysis Intelligence Indexcomposite · indexIndependentversion4.35−2.67vs Claude Fable 5.1obs. 11 Sept 2026artificialanalysis.aiT2
LiveBenchgeneral · global_averageOfficial boardreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−3.33 ptvs Claude Fable 5.125 Jun 2026livebench.aiT2
LiveBench Data Analysisgeneral · average score · variant=Data AnalysisOfficial boardvariantData Analysisreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−8.42 ptvs gpt-6-astra25 Jun 2026livebench.aiT2
LiveBench Languagegeneral · average score · variant=LanguageOfficial boardvariantLanguagereasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−2.00 ptvs Claude Fable 525 Jun 2026livebench.aiT2
LiveBench Instruction Followinginstruction-following · average score · variant=IFOfficial boardvariantIFreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−17.6 ptvs Gemini 3.8 Flash25 Jun 2026livebench.aiT2
Humanity's Last Examknowledge · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysis5−4.26 ptvs Claude Fable 5.1obs. 11 Sept 2026artificialanalysis.aiT2
LiveBench Mathematicsmath · average score · variant=MathematicsOfficial boardvariantMathematicsreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−1.28 ptvs Claude Fable 5.125 Jun 2026livebench.aiT2
MMMU-Promultimodal · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysis5−2.14 ptvs gpt-6-astraobs. 11 Sept 2026artificialanalysis.aiT2
GPQA Diamondreasoning · accuracy · variant=GPQA Diamond · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantGPQA Diamondreasoning_efforthighconditions differ across rows → partially comparable5−2.52 ptvs gpt-6-astraobs. 11 Sept 2026artificialanalysis.aiT2
LiveBench Reasoningreasoning · average score · variant=ReasoningOfficial boardvariantReasoningreasoning_effortmaxrelease2026-06-25conditions differ across rows → partially comparable1−1.44 ptvs gpt-6-astra25 Jun 2026livebench.aiT2

Current rows only, grouped by benchmark → canonical metric → comparability group (task configuration). Effort variants folded into this model appear as rows of the same group. 43 current rows in total. “vs leader” compares with the current leader of the benchmark's primary group only; other groups are not directly comparable. Comparability rules →

Providers & Pricing4

All offers in the price terminal →
Provider deployments, cheapest output first
ProviderContextInput / 1MCached inCache writeOutput / 1MNative unitsStatusObservedSource
Anthropic APIcheapest outputanthropic/claude-opus-5:batch1Mout 128K$0.25$3.13—active3 h agosince 11 Sept 2026openrouter.aiT2
Anthropic APIclaude-opus-5————active12 h agosince 11 Sept 2026anthropic.comT2
Anthropic API—$0.50$6.25cache_write_1h_per_mtok=10active12 h agosince 11 Sept 2026platform.claude.comT1
Anthropic APIanthropic/claude-opus-51Mout 128K$0.50$6.25—active3 h agosince 11 Sept 2026openrouter.aiT2

USD per 1M tokens as published by each provider; native units (per-request fees, flex/priority tiers) are kept verbatim. Rows are append-only — every price change is kept in the history below. Cost of a workload →

Price history

Step lines per provider; amber markers are recorded changes. Click a marker or a row for the evidence behind that price.

Output price · USD / 1M tokens 1 provider

Output price history of Claude Opus 5$0$10$20$30Sept 26Sept 26Sept 26Anthropic API: first observed → $25 · 11 Sept 2026Anthropic API: $25 → $12.5 · 11 Sept 2026
  • Anthropic API
  • Anthropic API$25 → $12.511 Sept 2026
  • Anthropic APIfirst observed $2511 Sept 2026

Input price · USD / 1M tokens 1 provider

Input price history of Claude Opus 5$0$2$4$6Sept 26Sept 26Sept 26Anthropic API: first observed → $5 · 11 Sept 2026Anthropic API: $5 → $2.5 · 11 Sept 2026
  • Anthropic API
  • Anthropic API$5 → $2.511 Sept 2026
  • Anthropic APIfirst observed $511 Sept 2026

Versions & Artifacts0

Version history

Context windowfirst observation only

11 Sept 2026current

Knowledge cutofffirst observation only

11 Sept 2026current

Max outputfirst observation only

11 Sept 2026current

Opennessfirst observation only

11 Sept 2026current

Statusfirst observation only

11 Sept 2026current

Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.

Artifacts 0

No artifact (checkpoint, quantisation, conversion or packaging) points to this model yet.

Change history45

Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.

Viewing AI Atlas as of 12 Aug 2026 — attributes exactly as the atlas knew them on that day; later corrections are not shown.

Back to today →

Claude Opus 5 was not yet in AI Atlas on 12 Aug 2026

First seen 11 Sept 2026. Nothing is inferred backwards: no attribute is shown for dates before the first observation.
45 claims · 36 properties9 conflicting

Familyfamily2conflicting claims

Claim history for Family
ValueValid from → toStatusSourceConfidenceExtractor
Opus → currentconflictingClaude Developer Platform docsT2mediumllm
Claude → currentcurrentClaude Developer Platform docsT1conflicteddeterministic

Versionversion1

Claim history for Version
ValueValid from → toStatusSourceConfidenceExtractor
5 → currentcurrentClaude Developer Platform docsT2mediumllm

Release daterelease_date1

Claim history for Release date
ValueValid from → toStatusSourceConfidenceExtractor
24 Jul 2026 → currentcurrentClaude Developer Platform docsT1highdeterministic

Statusstatus1

Claim history for Status
ValueValid from → toStatusSourceConfidenceExtractor
active → currentcurrentClaude Developer Platform docsT1highdeterministic

Opennessopenness1

Claim history for Openness
ValueValid from → toStatusSourceConfidenceExtractor
proprietary → currentcurrentClaude Developer Platform docsT1highdeterministic

Context windowcontext_length1

Claim history for Context window
ValueValid from → toStatusSourceConfidenceExtractor
1M tokens → currentcurrentClaude Developer Platform docsT1highdeterministic

Max outputmax_output_tokens1

Claim history for Max output
ValueValid from → toStatusSourceConfidenceExtractor
128K tokens → currentcurrentClaude Developer Platform docsT1highdeterministic

Knowledge cutoffknowledge_cutoff1

Claim history for Knowledge cutoff
ValueValid from → toStatusSourceConfidenceExtractor
May 2026 → currentcurrentClaude Developer Platform docsT1highdeterministic

Training data cutofftraining_data_cutoff1

Claim history for Training data cutoff
ValueValid from → toStatusSourceConfidenceExtractor
May 2026 → currentcurrentClaude Developer Platform docsT1highdeterministic

Modalitiesmodalities5conflicting claims

Claim history for Modalities
ValueValid from → toStatusSourceConfidenceExtractor
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
text, image → currentcurrentClaude Developer Platform docsT1conflicteddeterministic

Input modalitiesmodalities_input5conflicting claims

Claim history for Input modalities
ValueValid from → toStatusSourceConfidenceExtractor
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
image, text → currentconflictingOpenRouter public model & pricing listingT2mediumdeterministic
text, image → currentcurrentClaude Developer Platform docsT1conflicteddeterministic

Output modalitiesmodalities_output1

Claim history for Output modalities
ValueValid from → toStatusSourceConfidenceExtractor
text → currentcurrentClaude Developer Platform docsT1highdeterministic

Tokenizertokenizer1

Claim history for Tokenizer
ValueValid from → toStatusSourceConfidenceExtractor
Claude → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

API model idapi_model_id1

Claim history for API model id
ValueValid from → toStatusSourceConfidenceExtractor
claude-opus-5 → currentcurrentClaude Developer Platform docsT1highdeterministic

API aliasapi_alias1

Claim history for API alias
ValueValid from → toStatusSourceConfidenceExtractor
claude-opus-5 → currentcurrentClaude Developer Platform docsT1highdeterministic

Retirement dateretirement_date1

Claim history for Retirement date
ValueValid from → toStatusSourceConfidenceExtractor
24 Jul 2027 → currentcurrentClaude Developer Platform docsT1highdeterministic

Retirement tentativeretirement_tentative1

Claim history for Retirement tentative
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentClaude Developer Platform docsT1highdeterministic

Official pageofficial_url1

Claim history for Official page
ValueValid from → toStatusSourceConfidenceExtractor
https://platform.claude.com/docs/en/models/opus-5/overview → currentcurrentClaude Developer Platform docsT1highdeterministic

Bedrock model idbedrock_model_id1

Claim history for Bedrock model id
ValueValid from → toStatusSourceConfidenceExtractor
anthropic.claude-opus-5 → currentcurrentClaude Developer Platform docsT1highdeterministic

Default effortdefault_effort1

Claim history for Default effort
ValueValid from → toStatusSourceConfidenceExtractor
high → currentcurrentClaude Developer Platform docsT1highdeterministic

Descriptiondescription1

Claim history for Description
ValueValid from → toStatusSourceConfidenceExtractor
For complex agentic coding and enterprise work → currentcurrentClaude Developer Platform docsT1highdeterministic

File inputfile_input1

Claim history for File input
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Foundry model idfoundry_model_id1

Claim history for Foundry model id
ValueValid from → toStatusSourceConfidenceExtractor
claude-opus-5 → currentcurrentClaude Developer Platform docsT1highdeterministic

Latency tierlatency_tier1

Claim history for Latency tier
ValueValid from → toStatusSourceConfidenceExtractor
Moderate → currentcurrentClaude Developer Platform docsT1highdeterministic

Aa median output tokens per secondmetric.aa_median_output_tokens_per_second1

Claim history for Aa median output tokens per second
ValueValid from → toStatusSourceConfidenceExtractor
51.5 → currentcurrentArtificial AnalysisT2mediumdeterministic

Openrouter idopenrouter_id1

Claim history for Openrouter id
ValueValid from → toStatusSourceConfidenceExtractor
anthropic/claude-opus-5 → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Openrouter listed atopenrouter_listed_at1

Claim history for Openrouter listed at
ValueValid from → toStatusSourceConfidenceExtractor
24 Jul 2026 → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Reasoningreasoning1

Claim history for Reasoning
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Safety notessafety_notes1

Claim history for Safety notes
ValueValid from → toStatusSourceConfidenceExtractor
Opus 5 does not advance the frontier in risky, dual-use capabilities. It remains behind Mythos 5 in biology research and offensive cybersecurity. Cyber classifiers block binary-based vulnerability scanning, penetration testing, and exploit generation. Requests flagged by safety classifiers fall back to Opus 4.8 by default. The Cyber Verification Program (CVP) provides access to a version with fewer security restrictions. → currentcurrentClaude Developer Platform docsT2mediumllm

Structured outputstructured_output1

Claim history for Structured output
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Supported parameterssupported_parameters1

Claim history for Supported parameters
ValueValid from → toStatusSourceConfidenceExtractor
include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, stop, structured_outputs, temperature, tool_choice, tools, verbosity → currentcurrentOpenRouter public model & pricing listingT2mediumdeterministic

Thinkingthinking1

Claim history for Thinking
ValueValid from → toStatusSourceConfidenceExtractor
Adaptive → currentcurrentClaude Developer Platform docsT1highdeterministic

Tool callingtool_calling1

Claim history for Tool calling
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentClaude Developer Platform docsT1highdeterministic

Training data notestraining_data_notes1

Claim history for Training data notes
ValueValid from → toStatusSourceConfidenceExtractor
Intentionally avoided training on cyber tasks. Improved on cyber tasks as a result of becoming more generally capable. → currentcurrentClaude Developer Platform docsT2mediumllm

Vertex model idvertex_model_id1

Claim history for Vertex model id
ValueValid from → toStatusSourceConfidenceExtractor
claude-opus-5 → currentcurrentClaude Developer Platform docsT1highdeterministic

Visionvision1

Claim history for Vision
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentClaude Developer Platform docsT1highdeterministic

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

Provenance

Attributed facts

37

Source tiers

T1T225 / 12

Freshest observation

2 h ago

Conflicts

9 flagged

Source documents 7

Source documents
SourceDocumentTypeTierLast observedSnapshots
Claude Developer Platform docsdocs.claude.com/en/docs/about-claude/pricing.md pricingT1· Official4 h ago1
Claude Developer Platform docsdocs.claude.com/en/docs/about-claude/models/overview.md model_docsT1· Official4 h ago1
Claude Developer Platform docsanthropic.com/news/claude-opus-5 newsT1· Official12 h ago1
Claude Developer Platform docsplatform.claude.com/docs/en/models/opus-5/overview.md model_pageT1· Official12 h ago1
OpenRouter public model & pricing listingopenrouter.ai/api/v1/models catalogueT2· Quality secondary3 h ago7
LiveBenchlivebench.ai/table_2026_06_25.csv leaderboardT2· Quality secondary10 h ago1
Artificial Analysisartificialanalysis.ai/leaderboards/models leaderboardT2· Quality secondary10 h ago1

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.

Data quality (77/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →