Claude API Pricing Calculator & Complete Cost Guide
Calculate Claude API costs across the current Opus, Sonnet, and Haiku lineup per token and per month, with batch and prompt-caching discounts.
Pricing TLDR
- • $5 free credits for new users (no credit card required)
- • Pay-per-token: Haiku 5.5 (from $0.10/$0.50) · Sonnet 5.5 ($2/$10) · Opus 5.5 ($4/$20) per 1M tokens
- • 50% batch discount · up to 90% prompt-caching savings · 1M context on Opus 5.5/5/4.8/4.7 and Sonnet 5.5/5/4.6 at standard pricing
Claude API Cost Calculator: Monthly Pricing
Calculate by
Input Tokens
Output Tokens
API Calls / Month
Quick Examples:
Cost Optimization:
Claude Opus 5.5 (claude-opus-5-5)
Context
Quality
Per 1M Tokens
In: $4.00
Out: $20.00
Monthly Cost
Claude Sonnet 5.5 (claude-sonnet-5-5)
Context
Quality
Per 1M Tokens
In: $2.00
Out: $10.00
Monthly Cost
Claude Fable 5.1 (claude-fable-5-1)
Context
Quality
Per 1M Tokens
In: $10.00
Out: $50.00
Monthly Cost
Claude Opus 5 (claude-opus-5)
Context
Quality
Per 1M Tokens
In: $5.00
Out: $25.00
Monthly Cost
Claude Fable 5 (claude-fable-5)
Context
Quality
Per 1M Tokens
In: $10.00
Out: $50.00
Monthly Cost
Claude Haiku 5.5 (claude-haiku-5-5)
Context
Quality
Per 1M Tokens
In: $0.10
Out: $0.50
Monthly Cost
Claude Opus 4.8 (claude-opus-4-8)
Context
Quality
Per 1M Tokens
In: $5.00
Out: $25.00
Monthly Cost
Claude Opus 4.7 (claude-opus-4-7)
Context
Quality
Per 1M Tokens
In: $5.00
Out: $25.00
Monthly Cost
Claude Sonnet 5 (claude-sonnet-5)
Context
Quality
Per 1M Tokens
In: $2.00
Out: $10.00
Monthly Cost
Claude Opus 4.6 (claude-opus-4-6)
Context
Quality
Per 1M Tokens
In: $5.00
Out: $25.00
Monthly Cost
Claude Sonnet 4.6 (claude-sonnet-4-6)
Context
Quality
Per 1M Tokens
In: $3.00
Out: $15.00
Monthly Cost
Claude Opus 4.5 (claude-opus-4-5)
Context
Quality
Per 1M Tokens
In: $5.00
Out: $25.00
Monthly Cost
Claude Sonnet 4.5 (claude-sonnet-4-5)
Context
Quality
Per 1M Tokens
In: $3.00
Out: $15.00
Monthly Cost
Claude Haiku 4.5 (claude-haiku-4-5)
Context
Quality
Per 1M Tokens
In: $1.00
Out: $5.00
Monthly Cost
Claude Mythos 5.1 (claude-mythos-5-1)
Context
Quality
Per 1M Tokens
In: $10.00
Out: $50.00
Monthly Cost
Claude Mythos 5 (claude-mythos-5)
Context
Quality
Per 1M Tokens
In: $10.00
Out: $50.00
Monthly Cost
Tired of manually tracking your Claude credits?
Track your Claude API credit balance and spend in real-time.
Free 7-day trial. No sign-up, no credit card.
About Claude API
What is Claude API?
The Claude API provides programmatic access to Claude, a family of AI models created by Anthropic and designed for safety and helpfulness. Claude models (Opus, Sonnet, and Haiku) excel at tasks like content generation, code assistance, analysis, and conversation. Each model offers different capabilities and price points to match your specific needs.
- Multiple Model Tiers: Choose from Claude Fable 5.1 (premium reasoning), Opus 5.5 (current flagship), Opus 5, Opus 4.8, Opus 4.7, Opus 4.6, Opus 4.5, Fable 5, Mythos 5.1 and Mythos 5 (verified organizations only), Sonnet 5.5 (current mid-tier), Sonnet 5, Sonnet 4.6, Sonnet 4.5 (200K context, retiring Nov 30, 2026), Haiku 5.5 (speed-cost tier), and Haiku 4.5. Older models (Opus 4.1, Opus 4, Sonnet 4, Haiku 3.5, Haiku 3) have been deprecated or retired by Anthropic.
- Advanced Context Windows: All current Claude models support at least 200K tokens (~150K words). Claude Fable 5.1, Opus 5.5, 5, 4.8, 4.7, 4.6, Sonnet 5.5, 5, 4.6, Fable 5, and Mythos 5.1 and 5 support 1M context at standard pricing across the full window, perfect for analyzing large codebases or document collections. Haiku 5.5 also has 1M context but charges higher rates for prompts over 100K tokens.
- Token-Based Pricing: Pay only for what you use with separate pricing for input tokens (your prompts) and output tokens (Claude's responses). Output typically costs 5x more than input across the current lineup.
When to Use Claude API
Start with Claude Haiku 5.5 for cost-effective high-volume work, move to Sonnet 5.5 for general coding and reasoning, and reserve Opus 5.5 for the hardest tasks. Use Batch API and prompt caching to significantly reduce costs for production workloads.
Ideal for
- Customer support chatbots with natural conversation flow
- Code completion and review in development environments
- Large document summarization and analysis
- Creative content generation for marketing and blogs
- Automated data extraction from unstructured text
Not ideal for
- Real-time applications requiring <100ms latency
- Simple pattern matching tasks (use cheaper alternatives)
- Applications needing guaranteed deterministic outputs
- Tasks requiring specialized domain knowledge without context
Claude API Pricing Breakdown
Free Tier
New users receive $5 in free credits with no credit card required. These credits don't expire and are automatically applied to your usage across all Claude models.
- Sign up at platform.claude.com, no credit card required
- Receive $5 free credits instantly upon registration
- Credits never expire and work across all models
- Generate API key and start building immediately
Cost Optimization Features
Batch API (50% Discount)
Process non-urgent workloads within 24 hours at half price. Example: Opus 4.7 drops to $2.50/$12.50 and Sonnet 4.6 drops to $1.50/$7.50 per million tokens. Perfect for data processing and content generation.
Prompt Caching (up to 90% Savings)
Cache frequently used prompts, system messages, or documents. Two write tiers: 5-minute TTL at 1.25x base price, and 1-hour TTL at 2x base price. Cache hits cost at most 0.1x the base input price, and less on the newest models. Savings compound on repeated requests.
Prepaid Credits System
Buy credits upfront with no monthly fees. Set auto-reload thresholds to never run out of credits, and configure spending limits in the Console.
Real-time Usage Tracking
Monitor usage and costs in real-time. Set spending limits and alerts. Failed requests aren't charged. Rate limits increase automatically with usage tier.
Claude API Monthly Cost Estimates
Light Use
$10-50/mo
• Personal projects
• <1K requests/day
• Haiku 5.5 recommended
Medium Use
$50-200/mo
• Small apps
• 1-5K requests/day
• Haiku 5.5 or Sonnet 5.5
Heavy Use
$200-1K/mo
• Production apps
• 5-20K requests/day
• Mixed models
Enterprise
$1K+/mo
• Large scale
• 20K+ requests/day
• Volume discounts
6 Claude API Cost Optimization Tips
Use Prompt Caching
Save 90% or more on repeated content by caching system prompts, instructions, or large documents. Two write tiers: 5-minute TTL at 1.25x base, 1-hour TTL at 2x base. Cache hits cost at most 0.1x the base input price, and less on the newest models. Example: 10K daily requests with 80% cache hits saves significantly on Sonnet 4.6.
Enable Batch API
Get 50% discount on all models by processing non-urgent workloads within 24 hours. Sonnet 4.6 drops to $1.50/$7.50 per M tokens; Opus 4.7 drops to $2.50/$12.50. Perfect for data processing, content generation, and analysis.
Start with Haiku 5.5
Use Claude Haiku 5.5 ($0.10/$0.50 per M tokens for prompts up to 100K, the cheapest current Claude) for classification, routing, and simple chat. Only escalate to Sonnet 5.5 or Opus 5.5 when complexity warrants it.
Implement Fallback Chains
Create smart escalation: Haiku 5.5 for simple tasks → Sonnet 5.5 for general coding and reasoning → Opus 5.5 for the hardest agentic or long-context work. This ensures you never overpay for simple queries.
Set Spending Limits
Configure auto-reload thresholds and maximum spending limits in your Console. Set up alerts to notify you before reaching budget limits to prevent unexpected charges.
Monitor Claude API Token Usage in Real-Time
Track Claude API spending per model with CostGoat's token-level visibility. Get instant alerts when switching from Haiku to Opus models, when prompt caching savings drop unexpectedly, or when batch processing opportunities are missed.
Claude Model Selection Guide
Use Case
Customer Support Chat
Recommended Model
Claude Haiku 5.5
Cheapest current Claude
Monthly Cost (Est.)
~$3-12
Why This Model?
Lowest cost for simple queries ($0.10/$0.50 per 1M)
Use Case
Code Generation
Recommended Model
Claude Sonnet 5.5
Current Sonnet
Monthly Cost (Est.)
~$100-500
Why This Model?
Strong coding, 1M context at standard pricing
Use Case
Research & Analysis
Recommended Model
Claude Opus 5.5
Current Opus flagship
Monthly Cost (Est.)
~$150-700
Why This Model?
Top-tier reasoning, handles complex multi-step tasks
Use Case
Content Writing
Recommended Model
Claude Sonnet 4.6
Balanced
Monthly Cost (Est.)
~$75-300
Why This Model?
Balanced performance and cost for quality content
Use Case
Data Extraction
Recommended Model
Claude Haiku 5.5 + Batch
Best value
Monthly Cost (Est.)
~$2-6
Why This Model?
Cheapest option ($0.05/$0.25 with batch discount)
Use Case
Legal/Medical Review
Recommended Model
Claude Opus 5.5
Current Opus flagship
Monthly Cost (Est.)
~$150-700
Why This Model?
Top accuracy, superior domain-specific reasoning
Claude API Rate Limits & Tiers
Usage Tier
Monthly Spend Cap
Claude Fable 5.x
1K RPM
Opus, Sonnet, Haiku
1K RPM
Usage Tier
Monthly Spend Cap
Claude Fable 5.x
2K RPM
Opus, Sonnet, Haiku
5K RPM
Usage Tier
Monthly Spend Cap
Claude Fable 5.x
4K RPM
Opus, Sonnet, Haiku
10K RPM
Usage Tier
Monthly Spend Cap
Claude Fable 5.x
ITPM / OTPM: Input / Output Tokens Per Minute | RPM: Requests Per Minute. Anthropic places organizations on a tier automatically based on usage history, and new organizations may start in a lower Evaluation tier. Cached input tokens do not count toward ITPM on current models. The Fable limit is shared by Fable 5.1 and Fable 5 (Mythos 5.1 and 5 share a separate one), and Opus 4.8, 4.7, 4.6, and 4.5 share one Opus 4.x limit; Opus 5.5, Opus 5, Sonnet 5.5, and Sonnet 5 each have their own.
Start Tracking Your Claude API Spending
Monitor your Claude API credit balance and agent quota resets from your menubar. Alerts before spending spikes.
Free 7-day trial. No sign-up, no credit card.
Claude API Pricing FAQ
Common questions about Claude API costs, billing, and optimization
Pricing per 1M tokens (input / output). Fable tier: Claude Fable 5.1 and Fable 5 at $10 / $50 with 1M context; Mythos 5.1 and Mythos 5, for verified organizations only, match them on price. Opus tier: Claude Opus 5.5 (flagship) at $4 / $20 with 1M context; Opus 5, 4.8, and 4.7 at $5 / $25 with 1M context at standard pricing. Older Opus 4.6 same rate; Opus 4.5 at 200K context. Sonnet tier: Claude Sonnet 5.5 and Sonnet 5 at $2 / $10 with 1M context; Sonnet 4.6 at $3 / $15; Sonnet 4.5 at a 200K context window (retiring Nov 30, 2026). Haiku tier: Claude Haiku 5.5 at $0.10 / $0.50 for prompts up to 100K tokens and $0.50 / $2.50 above that, with 1M context; Haiku 4.5 at $1 / $5. Batch API applies a 50% discount; prompt caching saves up to 90% on cached input.
Opus 5.5 is Anthropic's flagship, best for the hardest reasoning, complex coding, and agentic tasks ($4/$20). Sonnet 5.5 is the mid-tier sweet spot, strong coding and reasoning at much lower cost ($2/$10). Haiku 5.5 is the speed-cost tier, best for classification, routing, and simple chat ($0.10/$0.50 for prompts up to 100K tokens). All offer prompt caching and Batch API for cost savings.
Depends on usage. Claude Pro ($20/month, or $17/month annual) gives heavy use of the chat UI with priority access, best if you mostly chat interactively. Claude Max starts at $100/month with 5x and 20x usage tiers for power users. The API has no monthly fee and charges per token, better for programmatic use, scripts, batch jobs, and apps integrating Claude. The API also exposes features (prompt caching, batch processing, fine-grained model selection) that aren't available via the subscription plans.
Claude is typically more expensive per token than competitors at similar quality tiers. Opus 5.5 ($4/$20) sits at the frontier alongside the top OpenAI and Google models, at a price premium over Google's flagship. Sonnet 5.5 ($2/$10) costs more than the comparable Gemini Flash tier. Claude's strengths are agentic coding, long-context recall, and instruction-following, worth the premium when those matter.
New users get $5 in free credits when they sign up, no credit card required. At current rates that's enough for roughly 50M input tokens on Haiku 5.5, 2.5M on Sonnet 5.5, or 1.25M on Opus 5.5. Credits don't expire and are applied automatically. Beyond the initial credits there's no permanent free tier, so all usage is billed.
