Anthropic / lineage

Anthropic Claude family

Opus, Sonnet, Haiku and Fable are distinct named rows. Direct API retirement dates, partner-platform schedules and restricted model access require separate checks.

Family research reviewed · Official source notes below

Standard models
15
Independently rated
14 / 15
Known token rates
15 / 15
Global balanced frontier
2

Before you choose

Retirement scope

Anthropic’s lifecycle page applies to its own platforms and points to separate Bedrock and Google Cloud schedules. A minimum availability commitment is not an announced retirement date.

Access and model identity

Check the exact model identity and access policy on the serving platform. Restricted Mythos access does not establish endpoint availability or transfer a Fable rating.

Compare exact catalogue rows

USD per 1M tokens. In / out are listed base rates; expand billing conditions before planning a request. LMArena scores measure the general preference board, not success on your engineering task. Unknown stays unknown.

Fable

ModelLMArenaInput / output $/MContext / max outputCapabilities & conditions
Claude Fable 5.1anthropic/claude-fable-5.11507.6±8.2 interval half-widthEffort: max$10.00 / $50.00Listed base rates
Billing details
Cached input
$0.250
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesWeights unknownReleased 2026-09-01
Claude Fable 5anthropic/claude-fable-51492.6±4.8 interval half-widthEffort: default$10.00 / $50.00Listed base rates
Billing details
Cached input
$1.00
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-06-09

Opus

ModelLMArenaInput / output $/MContext / max outputCapabilities & conditions
Claude Opus 5.5anthropic/claude-opus-5.5Unrated$4.00 / $20.00Listed base rates
Billing details
Cached input
$0.200
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesWeights unknownReleased 2026-09-22
Claude Opus 5anthropic/claude-opus-51504.9±4.2 interval half-widthEffort: high$5.00 / $25.00Listed base rates
Billing details
Cached input
$0.500
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-07-24
Claude Opus 4.8anthropic/claude-opus-4.81460.8±4.2 interval half-widthEffort: high$5.00 / $25.00Listed base rates
Billing details
Cached input
$0.500
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-05-27
Claude Opus 4.7anthropic/claude-opus-4.71490.0±3.9 interval half-widthEffort: high$5.00 / $25.00Listed base rates
Billing details
Cached input
$0.500
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-04-16
Claude Opus 4.6anthropic/claude-opus-4.61503.0±3.5 interval half-widthEffort: high$5.00 / $25.00Listed base rates
Billing details
Cached input
$0.500
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-02-04
Claude Opus 4.5anthropic/claude-opus-4.51450.5±3.2 interval half-widthEffort: default$5.00 / $25.00Listed base rates
Billing details
Cached input
$0.500
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

200,000Output 64,000file · image · textTools yes · reasoning yesClosed weightsReleased 2025-11-24
Claude Opus 4.1anthropic/claude-opus-4.11418.1±3.0 interval half-widthEffort: default$15.00 / $75.00Listed base rates
Billing details
Cached input
$1.50
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

200,000Output 32,000image · text · fileTools yes · reasoning yesWeights unknownReleased 2025-08-05

Sonnet

ModelLMArenaInput / output $/MContext / max outputCapabilities & conditions
Claude Sonnet 5anthropic/claude-sonnet-51442.2±4.6 interval half-widthEffort: high$2.00 / $10.00Listed base rates
Billing details
Cached input
$0.200
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-06-30
Claude Sonnet 4.6anthropic/claude-sonnet-4.61458.3±3.6 interval half-widthEffort: default$3.00 / $15.00Listed base rates
Billing details
Cached input
$0.300
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 128,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2026-02-17
Claude Sonnet 4.5anthropic/claude-sonnet-4.51438.3±2.9 interval half-widthEffort: default$3.00 / $15.00Context tiers apply
Billing details
Cached input
$0.300
Separate reasoning
Not separately listed
  • Above 200,000 input tokens: $6.00 in / $22.50 out; cached $0.600.

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

1,000,000Output 64,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2025-09-29
Claude Sonnet 4anthropic/claude-sonnet-41339.3±4.4 interval half-widthEffort: default$3.00 / $15.00Context tiers apply
Billing details
Cached input
$0.300
Separate reasoning
Not separately listed
  • Above 200,000 input tokens: $6.00 in / $22.50 out; cached $0.600.

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

200,000Output 64,000image · text · fileTools yes · reasoning yesWeights unknownReleased 2025-05-22

Haiku

ModelLMArenaInput / output $/MContext / max outputCapabilities & conditions
Claude Haiku 4.5anthropic/claude-haiku-4.51396.6±2.5 interval half-widthEffort: default$1.00 / $5.00Listed base rates
Billing details
Cached input
$0.100
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

200,000Output 64,000text · image · fileTools yes · reasoning yesClosed weightsReleased 2025-10-15
Claude 3 Haikuanthropic/claude-3-haiku1194.6±3.7 interval half-widthEffort: default$0.250 / $1.25Listed base rates
Billing details
Cached input
$0.030
Separate reasoning
Not separately listed

Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence.

200,000Output 4,096text · imageTools yes · reasoning noWeights unknownReleased 2024-03-13

Free and batch variants

These rows have different access or processing conditions. They are excluded from the standard-row counts above. A variant without a standard base remains a separately named variant.

Sources and scope

  • Claude model overview ↗

    Exact named models on the vendor platform; not a promise about every reseller

    Checked 2026-09-22 · web fetch returned 503; official URL fetched directly with HTTP 200 and inspected
  • Claude model deprecations ↗

    Claude API, Claude Platform on AWS and Microsoft Foundry; partner schedules differ

    Checked 2026-09-22 · official document opened and inspected
  • Claude Mythos ↗

    Restricted-access Mythos offering; not a public marketplace listing

    Checked 2026-09-22 · official page fetched with HTTP 200 and inspected after web open failed

Source review does not refresh a price or transplant a benchmark score. Read the method and correction policy.

Questions this page answers

Are the Anthropic Claude family models interchangeable?

No. These are exact catalogue rows with separate capability, pricing and rating evidence. Compare the checkpoint, serving endpoint, reasoning settings and required input types. A family name does not establish equivalent behaviour.

How many Anthropic rows are compared here?

15 standard rows, of which 14 have an accepted LMArena general score and 15 have known input and output token rates. 13 related free or batch variants are listed separately.

What do the prices and frontier counts mean?

Prices below are listed US dollars per million input or output tokens, before request-specific context tiers and other billing dimensions. The frontier count comes from the global balanced view. Neither a family price range nor a benchmark difference proves your task will be cheaper or more successful.

How current are the official notes?

The linked family research was reviewed 2026-09-22. That date is separate from each accepted catalogue observation and each vendor’s effective date. An official note applies to its named platform, not automatically to every reseller.

Evidence & Ask