Safety-first AI, built for real work.
Anthropic builds the Claude family — the leading models for coding, writing, and agentic tasks. Claude Opus 4.7 leads SWE-Bench Pro at 64.3%. Claude Sonnet 4.6 is the most widely used coding model in 2026.
Every Anthropic model in the directory, ranked by overall capability score.
Anthropic's latest generally available Opus model, tuned for frontier coding, AI agents, long-context reasoning, and high-fidelity vision.
Per 1 million tokens. Updated when providers change prices.
| Model | Input / 1M | Output / 1M | Context | Speed |
|---|---|---|---|---|
| Claude Opus 4.7 Premium | $5.00/1M | $25.00/1M | 1M | Deliberate |
| Claude Sonnet 4.6 Premium | $3.00/1M | $15.00/1M | 1M | Balanced |
| Claude Opus 4.6 Premium | $5.00/1M | $25.00/1M | 1M | Deliberate |
| Anthropic: Claude Sonnet 4 Balanced | $3.00/1M | $15.00/1M | 200K | Balanced |
| Anthropic: Claude Opus 4.5 |
Consumer plans for access without the API.
Head-to-head comparisons for the most-searched questions.
Newsletter
Pricing changes, new releases, and ranking shifts — straight to your inbox.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.
Claude Opus 4.7 is Anthropic's most capable model — it leads SWE-Bench Pro at 64.3% and handles the most complex coding and reasoning tasks. Claude Sonnet 4.6 is the best value at $3/1M input with 79.6% SWE-bench and 1M context.
Claude Opus 4.7 costs $5/1M input and $25/1M output. Claude Sonnet 4.6 is $3/$15. Claude Haiku is $0.80/$4 — the cheapest Claude option. Claude.ai Pro subscription is $20/month for consumer access.
Yes, by public benchmarks. Claude Opus 4.7 scores 64.3% on SWE-Bench Pro vs GPT-5.5 at 58.6%. Claude Sonnet 4.6 scores 79.6% on SWE-bench (classic). Both outperform GPT-5.4 on coding benchmarks. Claude also powers Cursor and Windsurf — the most popular AI coding editors.
Anthropic focuses on Constitutional AI — training Claude with principles rather than just reward signals. They have published their responsible scaling policy publicly and declined to release their most capable model (Claude Mythos) due to safety concerns.
The default model powering Cursor and Windsurf. 79.6% SWE-bench, 1M context window, and best-in-tier writing quality — all at $3/1M input.
Anthropic's most powerful model and the current leader on SWE-bench coding benchmarks with 80.8% — the strongest agentic coding model available.
Claude Sonnet 4 is Anthropic's mid-tier flagship model balancing strong reasoning, coding, and writing capabilities at a competitive price point. It sits between Haiku and Opus in Anthropic's lineup, offering substantive intelligence without the cost of top-tier models.
Claude Opus 4.5 is Anthropic's flagship reasoning and writing model, offering deep analytical capability and nuanced instruction-following across a 200K context window. It sits at the top of the Claude 4 lineup, prioritizing quality over speed.
Claude 3.5 Sonnet is Anthropic's mid-cycle flagship model, balancing strong reasoning, coding, and instruction-following with a 200K context window. It sits between Haiku and Opus in Anthropic's lineup, offering near-flagship quality at a lower cost than top-tier models.
Claude Haiku 4.5 is Anthropic's latest lightweight model in the Claude 4 family, optimized for speed and cost-efficiency while retaining strong instruction-following and reasoning capabilities. It supersedes Claude 4 Haiku with improved performance across coding, summarization, and conversational tasks.
Claude Sonnet 4.5 is Anthropic's mid-tier workhorse model, balancing strong reasoning and writing quality with reasonable latency at $3/$15 per million tokens. It slots above Haiku in capability while remaining more cost-accessible than Opus-tier models.
Claude Opus 4 is Anthropic's most capable flagship model, designed for complex reasoning, nuanced writing, and sophisticated multi-step tasks. It sits at the top of the Claude 4 family, prioritizing depth and quality over speed.
Fast and affordable Anthropic option that keeps writing quality surprisingly high for the price.
Claude Opus 4.1 is Anthropic's top-tier flagship model, designed for the most demanding tasks requiring deep reasoning, nuanced writing, and complex multi-step analysis. It sits at the apex of the Claude 4 family, prioritizing capability over cost and speed.
Claude 3.7 Sonnet with extended thinking enabled — Anthropic's hybrid reasoning model that explicitly deliberates before responding, surfacing its chain-of-thought for complex multi-step problems. It sits between standard Sonnet and full reasoning-only models, balancing depth with practical usability.
Claude 3.5 Haiku is Anthropic's fastest and most affordable model in the Claude 3.5 family, designed for high-throughput tasks requiring quick responses without sacrificing Claude's core instruction-following quality. It handles a massive 200K context window while maintaining speed suitable for production pipelines.
Claude 3 Haiku is Anthropic's fastest and most affordable Claude 3 model, designed for high-throughput tasks where speed and cost efficiency matter more than peak intelligence. It delivers surprisingly capable responses for a budget tier model, with a generous 200K context window.
| $5.00/1M |
| $25.00/1M |
| 200K |
| Deliberate |
| Anthropic: Claude 3.5 Sonnet Premium | $6.00/1M | $30.00/1M | 200K | Balanced |
| Anthropic: Claude Haiku 4.5 Balanced | $1.00/1M | $5.00/1M | 200K | Very fast |
| Anthropic: Claude Sonnet 4.5 Balanced | $3.00/1M | $15.00/1M | 1M | Balanced |
| Anthropic: Claude Opus 4 Premium | $15.00/1M | $75.00/1M | 200K | Deliberate |
| Claude 4 Haiku Budget | $0.80/1M | $4.00/1M | 200K | Very fast |
| Anthropic: Claude Opus 4.1 Premium | $15.00/1M | $75.00/1M | 200K | Deliberate |
| Anthropic: Claude 3.7 Sonnet (thinking) Balanced | $3.00/1M | $15.00/1M | 200K | Deliberate |
| Anthropic: Claude 3.5 Haiku Balanced | $0.80/1M | $4.00/1M | 200K | Very fast |
| Anthropic: Claude 3 Haiku Budget | $0.25/1M | $1.25/1M | 200K | Very fast |
