Claude Sonnet 4.5

Claude Sonnet 4.5 is engineered as a highly capable model focused on autonomy and complex problem-solving. It is designed to function independently for extended periods, maintaining coherence while orchestrating various tools and coordinating multiple agents. This architecture makes it particularly effective for long-horizon tasks, such as managing software development projects that span hours or days, or performing deep research that requires synthesizing large amounts of context into actionable deliverables.

Built with enhanced domain knowledge in areas like cybersecurity, finance, and software engineering, the model excels at proactive tasks such as patching vulnerabilities or conducting predictive financial analysis. Its design supports sophisticated agentic workflows, allowing it to handle complex tool interactions and maintain focus over long-running operations. By integrating seamlessly with development environments and productivity tools, it serves as a robust engine for users who need to automate intricate, multi-step processes that demand high levels of reasoning and reliability.

Vertex (Anthropic)claude-sonnet-4-5@20250929claude-sonnet

Open provider page Provider docs

Quick Info

Provider: Vertex (Anthropic)
Model key: claude-sonnet-4-5@20250929
Release date: Sep 29, 2025
Last updated: Sep 29, 2025
Knowledge cutoff: 2025-07-31
Input modalities
Output modalities
Capabilities

Cost

Input token cost: $3.00
Output token cost: $15.00

Limits

Output tokens: 64,000 tokens
Context window: 200,000 tokens

Transparent token rates

Compare Claude Sonnet 4.5 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Claude Sonnet 4.5

Vertex (Anthropic)

CoverageBenchmark

CodeRabbit's October 3, 2025 benchmark evaluated Claude Sonnet 4.5 against Sonnet 4 and Opus 4.1 across 25 difficult real-world pull requests containing known critical bugs, covering issues like concurrency, memory ordering, async race conditions, and API misuse. Results showed Sonnet 4.5 closes the gap with Opus 4.1 i The analysis characterized Sonnet 4.5 as delivering a pragmatic sweet spot for code review at scale, achieving near-Opus-level performance at significantly lower cost. However, the model exhibited a tendency toward hedging and cautious framing, sometimes producing conditional warnings that could complicate parsing for

Vertex (Anthropic)

CoverageBenchmark

Cirra AI's analyst-style technical breakdown covers Anthropic's Claude Sonnet 4.5, released September 29, 2025, positioning it as a frontier LLM targeting coding, agentic workflows, and computer use. Anthropic's own messaging quoted in the article calls it "the best model in the world for agents, coding, and computer u The analysis lists Sonnet 4.5's headline technical features: an expanded 200,000-token context window with up to 64K output, hybrid reasoning with extended thinking for multi-step tasks, and new platform tools including context editing, memory, checkpoints, and VS Code integration to support long-running agentic workfl