Key features

Agentic capability: designed to run multi-step agents and coordinate multi-agent workflows.
Coding & software use: Anthropic markets ==Claude Sonnet 4.5== as its best coding model, capable of end-to-end software work (design → implement → test → deploy).
Hybrid reasoning / Thinking mode: supports on-demand extended reasoning (“thinking”) to trade latency for higher accuracy on complex tasks.
Large context & output: advertised 200K context window and up to 64K output tokens for very large inputs/outputs.

Technical details (what the release changes)

Architecture & focus:==Claude Sonnet 4.5== is a Sonnet-series mid-to-frontier model tuned for multi-step tool use, extended task horizons, and improved deterministic behavior when interacting with external systems. Anthropic emphasizes improvements in computer use (automation of workflows across files, spreadsheets, and developer tools) and reasoning/math capabilities versus previous Sonnet releases.

Long-horizon operation: The release specifically highlights the model’s ability to run as a continuous agent for extended periods (reported examples up to ~30 hours of autonomous operation in internal and early-customer tests), a large step up from previous multi-hour limits. This is important for use cases that require continuous monitoring, orchestration, or multi-step software projects.

Tooling & context:==Claude Sonnet 4.5== ships with improved context-management and agent tooling (context editing, memory tools, multi-agent support), enabling developers to manage and persist agent state more robustly.

Performance benchmarks

SWE-bench Verified: 77.2% (200K thinking budget, scaffold + tools); 78.2% in 1M context; 82.0% reported for a “high-compute” candidate selection regime.
OSWorld (computer tasks): 61.4% for Sonnet 4.5 vs 42.2% for Sonnet 4 (four months earlier).
Autonomy length (internal tests): >30 hours continuous autonomous coding/agent operation (previous generation ~7 hours).
Operating-system/tool benchmark: Anthropic reports a jump to ~60% versus ~40% for the predecessor on an OS interaction benchmark — showing improved reliability when the model controls software.

Primary use cases

Software engineering & code generation: Large-scale code synthesis, multi-file project generation, autonomous coding agents.
Agentic automation & orchestration: Building long-lived agents that monitor systems, run repeated tasks, and coordinate tools (calendar, email, VMs, spreadsheets).
Data analysis & finance/cybersecurity workflows: Deep analysis over documents, litigation briefs, financial models, and cybersecurity triage where multi-step reasoning and reliable tool use matter.

How to access Claude Sonnet 4.5 API from CometAPI

Model version:

API call point	cursor suitable point
`claude-sonnet-4-5-20250929-thinking`	`cometapi-sonnet-4-5-20250929-thinking`
`claude-sonnet-4-5-20250929`	`cometapi-sonnet-4-5-20250929`
`claude-sonnet-4-5`	`cometapi-sonnet-4-5`

Claude Sonnet 4.5 API Pricing in CometAPI，20% off the official price:

Input Tokens: $2.4/ M tokens
Output Tokens: $12/ M tokens

Required Steps

Log in to cometapi.com. If you are not our user yet, please register first.
Sign into your CometAPI console.
Get the access credential API key of the interface. Click “Add Token” at the API token in the personal center, get the token key: sk-xxxxx and submit.

Use Method

Select the “ claude-sonnet-4-5-20250929-thinking ”or “ claude-sonnet-4-5-20250929","claude-sonnet-4-5 ” endpoint to send the API request and set the request body. The request method and request body are obtained from our website API doc. Our website also provides Apifox test for your convenience.
Replace <YOUR_API_KEY> with your actual CometAPI key from your account.
Insert your question or request into the content field—this is what the model will respond to.
. Process the API response to get the generated answer.

API Example

CometAPI provides a fully compatible REST API—for seamless migration. Key details:

Base URL: https://api.cometapi.com/v1/messages
Model Names: “ “ claude-sonnet-4-5-20250929-thinking ”or “ claude-sonnet-4-5-20250929","claude-sonnet-4-5 ”
Authentication: Bearer token via Authorization: Bearer YOUR_CometAPI_API_KEY header
Content-Type:application/json.

Yes, Claude Sonnet 4.5 excels at sustained long-horizon tasks, maintaining coherence across massive codebases during extended autonomous coding sessions.

Claude Sonnet 4.5 is optimized for agents with superior instruction following, tool selection, and self-correction. It achieved 61.4% on OSWorld for computer use tasks, making it particularly strong for browser automation and multi-step workflows.

Yes, Sonnet 4.5 is a hybrid reasoning model that can deliver instant responses or extended step-by-step thinking. API users have fine-grained control over thinking duration to balance performance with latency.

Claude Sonnet 4.5 supports up to 64,000 output tokens, which is particularly valuable for rich code generation, detailed planning documents, and long-form content.

Yes, Sonnet 4.5 has enhanced domain knowledge in finance and cybersecurity. It can monitor regulatory changes, perform predictive analysis, and generate creative attack scenarios for red teaming.

Choose Sonnet 4.5 for high-volume production workloads where you need a balance of advanced capabilities and cost efficiency. Use Opus 4.5 for mission-critical tasks requiring maximum intelligence where cost is secondary.

Yes, Sonnet 4.5 excels at computer use capabilities, reliably handling browser-based tasks from competitive analysis to procurement workflows to customer onboarding.

Version
claude-sonnet-4-5-20250929-thinking
claude-sonnet-4-5-20250929
claude-sonnet-4-5
cometapi-sonnet-4-5-20250929-thinking	cursor suitable point
cometapi-sonnet-4-5-20250929	cursor suitable point
cometapi-sonnet-4-5	cursor suitable point