Anthropic

Anthropic offers hosted Claude models for language workloads. The calculator prices the exact catalogued model from input and output tokens rather than a flat request average.

Categories
LLM
Models and endpoints
13
Guide reviewed
Oct 7, 2026

Models, translated into money

Choose a budget and compare what each model delivers. Every estimate includes its assumptions and pricing sources.

TaskBudget · USD
Compare13 models
Task
Budget USD

Claude Haiku 4.5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatLimited fit

Conversation quality and instruction following. Compare alternatives.

SupportGood fit

Instruction following and multi-turn conversations.

CodingLimited fit

Coding tasks and web development. Compare alternatives.

DocumentsLimited fit

Long queries and document work. Compare alternatives.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-haiku-4-5-standard · Current · Standard billing

Standard text
$1 input · $5 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-haiku-4-5-20251001
Tool calling
true
Structured output
true
context Window
200,000
max Output Tokens
64,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: 69 / 100 · limited fit

Support chat: 70 / 100 · good fit

Coding help: 47 / 100 · limited fit

Document work: 52 / 100 · limited fit

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Sonnet 5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatGood fit

Conversation quality and instruction following.

SupportGood fit

Instruction following and multi-turn conversations.

CodingGood fit

Coding tasks and web development.

DocumentsGood fit

Long queries and document work.

ResearchGood fit

Hard prompts, agents, and factual accuracy.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-sonnet-5-standard · Current · Standard billing

Standard text
$2 input · $10 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-sonnet-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: 88 / 100 · good fit

Support chat: 89 / 100 · good fit

Coding help: 83 / 100 · good fit

Document work: 84 / 100 · good fit

Research agent: 73 / 100 · good fit

Catalog 2026-10-08-7b05c4ce

Claude Opus 5.5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatStrong fit

Conversation quality and instruction following.

SupportStrong fit

Instruction following and multi-turn conversations.

CodingStrong fit

Coding tasks and web development.

DocumentsStrong fit

Long queries and document work.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-5-5-standard · Current · Standard billing

Standard text
$4 input · $20 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-5-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: 100 / 100 · strong fit

Support chat: 100 / 100 · strong fit

Coding help: 100 / 100 · strong fit

Document work: 100 / 100 · strong fit

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Fable 5.1

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatStrong fit

Conversation quality and instruction following.

SupportStrong fit

Instruction following and multi-turn conversations.

CodingStrong fit

Coding tasks and web development.

DocumentsStrong fit

Long queries and document work.

ResearchStrong fit

Hard prompts, agents, and factual accuracy.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-fable-5-1-standard · Current · Standard billing

Standard text
$10 input · $50 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-fable-5-1
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: 98 / 100 · strong fit

Support chat: 97 / 100 · strong fit

Coding help: 97 / 100 · strong fit

Document work: 96 / 100 · strong fit

Research agent: 99 / 100 · strong fit

Catalog 2026-10-08-7b05c4ce

Claude Sonnet 5.5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-sonnet-5-5-standard · Current · Standard billing

Standard text
$2 input · $10 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-sonnet-5-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Haiku 5.5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-haiku-5-5-standard · Current · Standard billing

Standard text
$0.10 input · $0.50 output / 1M tokens
Above 100,000 input tokens per request
$0.50 input · $2.50 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-haiku-5-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Opus 5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-5-standard · Current · Standard billing

Standard text
$5 input · $25 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Opus 4.8

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-4-8-standard · Current · Standard billing

Standard text
$5 input · $25 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-4-8
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Opus 4.7

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-4-7-standard · Current · Standard billing

Standard text
$5 input · $25 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-4-7
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Opus 4.6

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-4-6-standard · Current · Standard billing

Standard text
$5 input · $25 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-4-6
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Opus 4.5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-opus-4-5-standard · Current · Standard billing

Standard text
$5 input · $25 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-opus-4-5-20251101
Tool calling
true
Structured output
true
context Window
200,000
max Output Tokens
64,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Sonnet 4.6

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-sonnet-4-6-standard · Current · Standard billing

Standard text
$3 input · $15 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-sonnet-4-6
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

Claude Fable 5

Expired

Pricing is unavailable for this model.

Where it fits

Based on independent evidence
ChatNo data

Not enough current evidence to assess this task.

SupportNo data

Not enough current evidence to assess this task.

CodingNo data

Not enough current evidence to assess this task.

DocumentsNo data

Not enough current evidence to assess this task.

ResearchNo data

Not enough current evidence to assess this task.

How we calculated thisPricing unavailable
Typical chat assumption2,000 input · 500 output tokens per request

Official pricing

anthropic-claude-fable-5-standard · Current · Standard billing

Standard text
$10 input · $50 output / 1M tokens

Verified Oct 7, 2026

Model details

Model ID
claude-fable-5
Tool calling
true
Structured output
true
context Window
1,000,000
max Output Tokens
128,000
modality
text
processing Tier
standard
cache
No

Task evidence

General chat: No current score

Support chat: No current score

Coding help: No current score

Document work: No current score

Research agent: No current score

Catalog 2026-10-08-7b05c4ce

About Anthropic

Coverage, billing details, and what to check before you build.

Anthropic appears in the catalog as a language-model provider for applications that send text or structured context and receive generated output. The cost model is based on API usage, not access to a consumer chat product. Monthly spend therefore depends on how many model calls the application makes and how many input and output tokens each call uses.

Where Anthropic fits

A Claude model can provide the language layer for conversational products, document workflows, analysis, and research agents. In a larger stack, search and scraping remain separate services. Image and video generation are also separate categories. The calculator combines these categories only after each service has been priced with its own billing rule.

Chat and research workloads remain separate even when they use the same model. A short support message and a long research prompt can fall into different pricing or compatibility bands. Preserving both rows makes the estimate explainable and prevents a blended token average from selecting the wrong rate.

How the estimate is built

The catalogued Anthropic offers are priced with distinct input-token and output-token rates. For every workload, the engine multiplies call count by tokens per call, applies the appropriate rate, and sums the exact decimal results. Display rounding happens after the total is calculated.

Each model card above is generated from the pricing snapshot. It contains the exact API identifier, context and output limits when verified, current plan data, source links, and the oldest verification date that affects the estimate. Editorial copy intentionally does not duplicate those values, so a catalog refresh cannot leave a stale number inside the article.

Compatibility and evidence

Before locking a model, check whether the application requires tool calling, structured output, or a minimum context window. A capability marked unknown is handled conservatively. The offer will not be presented as compatible with a hard requirement until the underlying source supports that conclusion.

Fastest and Balanced recommendations need comparable evidence collected with the same host, workload, method, and recent time window. Provider marketing statements are not substituted for that evidence. When no comparable cohort exists, the cost remains available while those recommendations are clearly unavailable.

Limits of the budget

The estimate covers the catalogued API charge. It does not include taxes, application servers, storage, observability, retries, staff, or negotiated enterprise terms. Caching and batch pricing are outside the current MVP unless a dedicated offer models them later.

Use a model action above to open the calculator with Chat enabled and this offer locked. Adjust users, messages, and token sizes to match the planned product. Review the verification date and official source before relying on the result for procurement.