Skip to content

PRICING.md

Latest commit

 

History

History
76 lines (61 loc) · 3.16 KB

PRICING.md

File metadata and controls

76 lines (61 loc) · 3.16 KB

Embedded Pricing Reference

Source: Anthropic Pricing Last verified: 2026-03-22

This file is the source of truth for the embedded pricing table in the binary. Update this file and rebuild when Anthropic changes pricing.

Per-Model Rates ($ per million tokens)

Model ID Input Output 5m Cache Write 1h Cache Write Cache Read
claude-opus-4-6 5.00 25.00 6.25 10.00 0.50
claude-opus-4-5 5.00 25.00 6.25 10.00 0.50
claude-opus-4-5-20251101 5.00 25.00 6.25 10.00 0.50
claude-opus-4-1 15.00 75.00 18.75 30.00 1.50
claude-opus-4 15.00 75.00 18.75 30.00 1.50
claude-sonnet-4-6 3.00 15.00 3.75 6.00 0.30
claude-sonnet-4-5 3.00 15.00 3.75 6.00 0.30
claude-sonnet-4-5-20250929 3.00 15.00 3.75 6.00 0.30
claude-sonnet-4 3.00 15.00 3.75 6.00 0.30
claude-haiku-4-5 1.00 5.00 1.25 2.00 0.10
claude-haiku-4-5-20251001 1.00 5.00 1.25 2.00 0.10
claude-haiku-3-5 0.80 4.00 1.00 1.60 0.08

Pricing Multipliers

Modifier Multiplier Applies to Condition
5m cache write 1.25x input cache_creation (ephemeral_5m) Default cache tier
1h cache write 2.0x input cache_creation (ephemeral_1h) Extended cache
Cache read 0.1x input cache_read_input_tokens Cache hit
Fast mode 6.0x all input + output speed: "fast", Opus 4.6 only
Data residency 1.1x all input + output + cache inference_geo: US-only, Opus 4.6+
Long context 2.0x input, 1.5x output All tokens when input > 200k Sonnet 4.5/4 only
Batch API 0.5x all All token types Not applicable to Claude Code

Additional Costs

Feature Cost Field in JSONL
Web search $10 per 1,000 searches server_tool_use.web_search_requests
Web fetch Free (token cost only) server_tool_use.web_fetch_requests
Code execution Free with web search/fetch; $0.05/hr otherwise server_tool_use.code_execution_requests

How to Apply

cost = 0
for each unique request (deduplicated by requestId, final chunk):
    model = request.model
    rates = PRICING_TABLE[model]

    # Base costs
    input_cost   = request.input_tokens * rates.input / 1_000_000
    output_cost  = request.output_tokens * rates.output / 1_000_000

    # Cache costs (split by tier)
    cache_5m_cost = request.ephemeral_5m_input_tokens * rates.cache_write_5m / 1_000_000
    cache_1h_cost = request.ephemeral_1h_input_tokens * rates.cache_write_1h / 1_000_000
    cache_read_cost = request.cache_read_input_tokens * rates.cache_read / 1_000_000

    # Modifiers
    if request.speed == "fast":
        multiply all costs by 6.0
    if request.inference_geo == "us" and model >= opus-4-6:
        multiply all costs by 1.1
    if model in [sonnet-4-5, sonnet-4] and total_input > 200_000:
        input_cost *= 2.0
        output_cost *= 1.5

    # Server tool costs
    web_search_cost = request.web_search_requests * 10.0 / 1_000

    cost += input_cost + output_cost + cache_5m_cost + cache_1h_cost + cache_read_cost + web_search_cost