Mythos, Opus, Sonnet
Anthropic's large language models, named Mythos, Opus, and Sonnet, offer different capabilities and costs for various AI tasks.
Anthropic has developed several large language models, which are AI systems trained on vast amounts of text and code. They name these models Mythos, Opus, and Sonnet. Each model has different strengths and is suited for different kinds of tasks, like writing, coding, or analyzing information.
Think of them like different tools in a toolbox. Opus is generally their most capable model, suitable for complex tasks. Sonnet is a balance of capability and speed, good for many everyday uses. Mythos is their fastest model, designed for high-volume, lower-complexity tasks. Users can choose the model that best fits their needs and budget.
Anthropic's family of proprietary large language models includes Opus, Sonnet, and the newer Claude 3 models. Opus represents their highest-performing model, optimized for complex reasoning and analysis. Sonnet offers a balance between performance and speed, suitable for enterprise workloads. Mythos, previously referred to as Haiku, is their fastest and most cost-effective model, designed for near-instantaneous responses and high-throughput applications.
These models differ in their parameter count, training data, and architectural optimizations, leading to variations in capabilities such as context window length, reasoning ability, and latency. Anthropic provides access to these models via API, allowing developers to integrate them into applications and services, with pricing tiered according to model performance and usage.
Anthropic's use of 'Mythos', 'Opus', and 'Sonnet' highlights a growing trend in frontier AI labs to brand their specific model versions. This allows for clearer communication about the capabilities and intended uses of different AI systems, especially as models become more specialized.
- Introducing the Life Sciences Verification ProgramAnthropic · Anthropic News ·
- Anthropic Claude
- Anthropic uses 'Opus', 'Sonnet', and 'Mythos' to name its different large language models, each designed for distinct tasks and performance levels.
- Model Naming Conventions
- Other companies use different naming schemes for their
- Open Source Models
- Unlike open source models, Anthropic's models are proprietar
The names Mythos, Opus, and Sonnet are proprietary designations used by Anthropic for their family of large language models. These names reflect a deliberate branding strategy to differentiate their offerings.
Visit Anthropic's website and try interacting with their Claude chat interface. Experiment by asking the same complex question to different versions if available, or try a simple, fast query. Notice how the responses might differ in detail, speed, or style, reflecting the underlying model's design.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the difference between Opus and Sonnet?
- Opus is Anthropic's most capable model, excelling at complex tasks. Sonnet offers a balance of performance and speed, suitable for a wider range of everyday applications.
- Are these models good for coding?
- Yes, Anthropic's models, particularly the more capable ones like Opus, are designed to assist with coding tasks, including generation, explanation, and debugging.
First seen on Subvolts Sep 17, 2026. Updated Sep 24, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.