NezhaGateNezhaGate
简体中文English ✓Español한국어日本語
Log in Get API Key
Chat

Claude Opus 5

⧉
claude-opus-5

Claude Opus 5 is the next-generation flagship from Anthropic, with adaptive thinking on by default. It excels at complex reasoning, code engineering and agent orchestration, and supports image understanding, tool use and prompt caching. Fully OpenAI-API compatible: set model to claude-opus-5; or use the native Anthropic Messages API so Claude Code connects directly.

ChatDeep reasoningCodeImage understandingPrompt caching
Get API Key → API Docs

Live Test · Playground

Try out Claude Opus 5 right here (available after login).

Input

Advanced
Web search
Memory · multi-turn
Thinking

Conversation

Start a conversationType a message below to begin

About Claude Opus 5

Claude Opus 5 is Anthropic's next-generation flagship model. Adaptive thinking is on by default: the model decides how long to reason per request, answering simple questions directly and working through harder ones first, with the reasoning visible on the native API and in the playground. It is strong at complex reasoning, code engineering, agent orchestration and long-context documents, and natively supports image input, tool use and prompt caching. NezhaGate serves it two ways: the OpenAI-compatible endpoint (point base_url at NezhaGate, set model to claude-opus-5, reuse your OpenAI SDK unchanged), and the native Anthropic Messages API (point ANTHROPIC_BASE_URL at NezhaGate's /anthropic so Claude Code connects directly, preserving thinking, tool_use and prompt caching). Billing is pay-as-you-go and failed requests are not charged.

Use cases

Complex code engineering

Cross-file refactors, hard bug hunts, architecture review and migration plans - with tool use it can drive an agent through multi-step changes.

Deep reasoning and analysis

Multi-step logic, trade-off analysis, data interpretation and technology selection, with adaptive thinking spending more budget on the hard cases.

Long documents and codebases

Summarize, extract from and answer questions over contracts, papers, manuals and large repositories; prompt caching makes repeated questions much cheaper.

Agents and tool orchestration

Native tool_use supports function calling and multi-turn tool loops, which suits automated workflows and engineering assistants.

Mixed text and image tasks

Read charts, screenshots and design mockups and respond with analysis or code; both base64 and remote image URLs are accepted.

How to choose

Choose Claude Opus 5 when the task is genuinely hard and you want the model to think before it answers - complex reasoning, large code projects and agent orchestration are its home turf. For narrative, scripts or long-form writing, its sibling Claude Fable 5 has a finer touch; for cheaper everyday chat and coding, use Claude Sonnet 4.6; for million-token context, consider Gemini 3.1 Pro. All of them switch with a single model field on the same endpoint.

FAQ

What is Claude Opus 5?
Claude Opus 5 is Anthropic's next-generation flagship chat and coding model, with adaptive thinking on by default. It excels at complex reasoning, code engineering and agent orchestration, and supports image input, tool use and prompt caching. On NezhaGate you can call it through the OpenAI-compatible API or the native Anthropic Messages API.
Why is my temperature ignored?
Anthropic deprecated temperature / top_p / top_k on this generation, and sending them makes the upstream reject the call. So that existing OpenAI SDK code keeps working unchanged, the gateway strips those three parameters before forwarding: the request succeeds, no error is raised, and sampling runs at the model default. To steer the output, use reasoning_effort (low / medium / high / xhigh / max) to set thinking depth instead — all five rungs cost the same per token, deeper thinking simply emits more.
How do I see the thinking?
Call the native Anthropic Messages API and the response carries thinking content blocks; the NezhaGate playground shows the thinking summary directly. The OpenAI-compatible endpoint returns only the final answer, while thinking tokens are still billed as output.
Is prompt caching supported, and how is it billed?
Yes - cache writes and cache hits both verified working. Cache writes bill at the cache-write rate and hits at the cache-read rate, both listed on the pricing page. Long prompts may be cached upstream even without an explicit cache_control, which makes repeated calls noticeably cheaper.
Can Claude Code connect directly?
Yes. Point ANTHROPIC_BASE_URL at NezhaGate's /anthropic, use your gateway key as x-api-key and set model to claude-opus-5. Thinking, tool_use, prompt caching and streaming SSE all pass through unchanged.

Related Models

Explore other models you can integrate.

View all →
Claude Sonnet 4.6 Chat

Claude Sonnet 4.6

claude-sonnet-4-6

Balanced, efficient chat and coding model

$1.50/1M in · $7.50/1M out ↓50% View →
Claude Fable 5 Chat

Claude Fable 5

claude-fable-5

Anthropic Fable series · narrative and long-form writing

$8.00/1M in · $40.00/1M out View →
Gemini 3.1 Pro Chat

Gemini 3.1 Pro

gemini-3.1-pro

Next-generation flagship reasoning model from Google

$0.50/1M in · $3.00/1M out ↓75% View →
GPT-5.5 Chat

GPT-5.5

gpt-5.5

Flagship chat and reasoning model

$0.50/1M in · $3.00/1M out ↓90% View →
GPT-6 Astra Chat

GPT-6 Astra

gpt-6-astra

Next-generation OpenAI flagship - 1.05M context

$1.00/1M in · $5.00/1M out ↓90% View →
Gemini 3.6 Flash Chat

Gemini 3.6 Flash

gemini-3.6-flash

Next-generation fast thinking model · four thinking budgets

$0.60/1M in · $3.60/1M out View →