Anthropic's most powerful model revealed. SWE-bench Pro 77.8% (+24.4 over Opus). Currently restricted to Project Glasswing partners.
Anthropic Claude Models
The complete Anthropic Claude model lineup - 32 models spanning the Opus, Sonnet, and Haiku families. Compare every Claude model by score, pricing, context window, and capabilities. Updated hourly from live API data.
All Claude Models - Ranked by Score
Claude Pricing Comparison - Sorted by Output Price
Understanding the Claude Model Family
Claude Family Overview
Anthropic's Claude model family spans three tiers: Opus (premium), Sonnet (balanced), and Haiku (fast and affordable). Each generation brings improvements in reasoning, coding, instruction-following, and safety. The Claude 3.5 series represents the latest publicly available generation.
Opus vs Sonnet vs Haiku
Opus is the most capable tier, designed for complex reasoning and research. Sonnet offers the best balance of intelligence and speed for production workloads. Haiku is optimized for high-throughput, low-latency tasks at the lowest cost per token.
Use Case Guide
Use Opus for difficult analysis, long-form content, and complex code generation. Sonnet excels at everyday coding assistance, document summarization, and chat applications. Haiku is ideal for classification, content moderation, customer support bots, and high-volume API workloads where cost efficiency matters most.
API Access
All Claude models are available through Anthropic's direct API, as well as through Amazon Bedrock, Google Cloud Vertex AI, and other major cloud providers. Pricing varies by provider - the prices shown here reflect current per-token API rates, updated hourly.
相关页面
Explore Claude comparisons, rankings, and pricing across the full model landscape.
Anthropic offers Claude 3.5 Sonnet, Claude 3.5 Haiku, and Claude 3 Opus through their API. Claude 3.5 Sonnet is the most popular, offering excellent coding and reasoning at competitive prices.
Claude and GPT-4o are closely matched. Claude tends to excel at longer outputs, coding, and following nuanced instructions. GPT-4o has advantages in breadth of capabilities and tool ecosystem. Check our comparison for the latest benchmarks.
Claude 3.5 Sonnet costs $3/M input and $15/M output tokens. Claude 3.5 Haiku starts at $0.80/M input tokens for a more affordable option. Anthropic offers a free tier with rate limits.