Claude Sonnet 4

Unleash the power of Claude Sonnet 4, a cutting-edge AI model designed for exceptional coding, advanced reasoning, and precise instruction following.

Platform: Replicate

Reasoning AICoding AssistantInstruction FollowingText Generation

99.8k runs

Input: $3 per million tokens; Output: $15 per million tokens

Commercial

🚀Function Overview

A hybrid reasoning model providing near-instant responses and extended thinking capabilities for coding, analysis, and complex instruction following.

Key Features

Dual operating modes (standard vs. extended thinking)
State-of-the-art coding performance (72.7% on SWE-bench)
Enhanced instruction following and steerability
Parallel tool execution
Web search integration during extended thinking
65% reduction in shortcut/loophole behavior

Use Cases

•Coding tasks and development workflows
•Complex instruction following
•Multi-file codebase operations
•Autonomous application development
•Long-form reasoning and analysis
•Agent-based workflows

⚙️Input Parameters

prompt

string

Input prompt

image

string

Optional input image. Images are priced as (width px * height px)/750 input tokens

system_prompt

string

System prompt

max_tokens

integer

Maximum number of output tokens

max_image_resolution

number

Maximum image resolution in megapixels. Scales down image before sending it to Claude, to save time and money.

extended_thinking

boolean

Whether to enable extended thinking mode (only supported for Sonnet 3.7 and Sonnet 4)

thinking_budget_tokens

integer

Maximum number of tokens to use for extended thinking when enabled (only supported for Sonnet 3.7 and Sonnet 4)

💡Usage Examples

Example 1

Input Parameters

{
  "prompt": "Give me a recipe for tasty smashed avocado on sourdough toast that could feed all of California.",
  "max_tokens": 8192,
  "system_prompt": "You are a helpful assistant",
  "max_image_resolution": 0.5
}

Output Results

Here's a recipe to

feed all of California (about 39

million people)!

🥑 ## California

-Sized Smashed Avoca

do Toast ### Ingredients: - **

780 million avocados** (

20 per person - hey, it's California!)

- **39 million slices**

of sourdough bread - **39

million limes** (ju

iced) - **19.

5 million red onions** (f

inely diced) - **

390,000 pounds** of sea

salt - **195,000 pounds** of

black pepper - **39

million Roma tomatoes** (

diced, optional) - **19

.5 million jalapeños** (min

ced, for the brave)

### Instructions: 1.

Recruit every resident to help prep

(mandatory participation law pending) 2. Use

the entire Central Valley as your prep

kitchen 3. Mash av

ocados in swimming pools (cleane

d first, obviously) 4

. Toast bread using solar power from

the Mojave Desert 5

. Assembly line from San Diego to the

Oregon border 6.

Distribute via a fleet of food trucks,

drones, and very patient

pelicans ###

Serving suggestion: Serve on

the beaches simultaneously at 12

PM PT for the ultimate California experience! **

Estimated cost:** Approximately the

GDP of a small country **Prep

time:** 6 months with

full state cooperation **Clean

-up time:** We don't talk

about clean-up time

*Note: May cause a

global avocado shortage an

d significantly impact the housing market (

even more expensive now).*

Quick Actions

Use NowView Documentation

Technical Specifications

Hardware Type
Run Count: 99.8k
Commercial Use: Supported
Pricing: Input: $3 per million tokens; Output: $15 per million tokens
Platform: Replicate

Related Keywords

Coding AssistantReasoning ModelInstruction FollowingComplex AnalysisAgent-based WorkflowsLong-form ReasoningAutonomous Application Development

Related Models

OpenAI O1-Mini Language Model

A small model alternative to o1

OpenAI o4-mini Model

OpenAI's fast, lightweight reasoning model

Meta Llama 4 Scout Instruct

A 17 billion parameter model with 16 experts