OpenRouter 模型 ↗ · 2026-09-23
anthropic/claude-opus-5.5
Model release OpenRouter 模型 Source language: Chinese
OpenRouter 于 2026-09-23 收录该模型(收录时间不等于模型正式发布日期);上下文:1,000,000 tokens;输入价:$4.0000/百万 tokens;输出价:$20.0000/百万 tokens;定位(来源描述):Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon…
Full text
Anthropic: Claude Opus 5.5
anthropic/claude-opus-5.5
Compare
Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code review and bug finding, financial and scientific analysis, and reading dense charts, diagrams, and screenshots, and it is more careful than its predecessor about only stating figures and citing sources it can back up.
The model completes comparable tasks in fewer steps and with fewer tokens than Opus 5, and reports on its work in plainer language, with clear updates on what it did, what it found, and what it needs from the user. Thinking is always adaptive, so effort is the main lever for trading off depth, latency, and cost, and lower effort settings remain effective for latency-sensitive workloads.
Modalities
In / out price
$4 / $20per 1M
Context
1.0M
Released
Sep 22, 2026
Compare
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
Latency / throughput
$4.00$20.00$0.202.97s78 tok/s
100.00%
$4.00$20.00$0.202.59s82 tok/s
100.00%
$4.00$20.00$0.203.01s80 tok/s
100.00%
$4.00$20.00$0.204.63s71 tok/s
100.00%
$4.00$20.00$0.202.46s81 tok/s
100.00%
Not used in Standard routing:Why these endpoints are not used
$4.40$22.00$0.221.50s39 tok/s
100.00%
$4.40$22.00$0.223.96s61 tok/s
100.00%
$4.40$22.00$0.22----
100.00%
$4.40$22.00$0.223.45s84 tok/s
100.00%
$4.40$22.00$0.223.15s78 tok/s
100.00%
Fast
$8.00$40.00$0.401.04s175 tok/s
100.00%
ProviderInput / Output /M
Input $4.00 / Output $20.00
78 tok/s100.00% uptime
Input $4.00 / Output $20.00
82 tok/s100.00% uptime
Input $4.00 / Output $20.00
80 tok/s100.00% uptime
Input $4.00 / Output $20.00
71 tok/s100.00% uptime
Input $4.00 / Output $20.00
81 tok/s100.00% uptime
Not used in Standard routing:Why these endpoints are not used
Input $4.40 / Output $22.00
39 tok/s100.00% uptime
Input $4.40 / Output $22.00
61 tok/s100.00% uptime
Input $4.40 / Output $22.00
-- tok/s100.00% uptime
Input $4.40 / Output $22.00
84 tok/s100.00% uptime
Input $4.40 / Output $22.00
78 tok/s100.00% uptime
Input $8.00 / Output $40.00
Fast
175 tok/s100.00% uptime
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
Input
$0.8089/M tokens
Output
$20.27/M tokens
Claude Opus 5.5 — Price history
Price history
No effective price data for this selection
$0.720$20.0088.3%22.4%
$0.779$20.0086.8%36.5%
$0.812$20.0086.5%0.3%
$0.868$20.0085.4%4.3%
$0.989$20.0085.1%29.0%
Not used in Standard routing:Why these endpoints are not used
$0.285$22.0099.2%4.0%
$0.645$22.0091.9%2.1%
$0.672$22.0091.7%0.6%
Fast
$0.877$40.0094.9%1.0%
$0.937$22.0085.3%0.0%
$4.399$22.000.0%0.0%
ProviderInput / Output /M
Token share
Effective in $0.720 / Effective out $20.00
Token share
22.4%
Effective in $0.779 / Effective out $20.00
Token share
36.5%
Effective in $0.812 / Effective out $20.00
Token share
0.3%
Effective in $0.868 / Effective out $20.00
Token share
4.3%
Effective in $0.989 / Effective out $20.00
Token share
29.0%
Not used in Standard routing:Why these endpoints are not used
Effective in $0.285 / Effective out $22.00
Token share
4.0%
Effective in $0.645 / Effective out $22.00
Token share
2.1%
Effective in $0.672 / Effective out $22.00
Token share
0.6%
Effective in $0.877 / Effective out $40.00
FastToken share
1.0%
Effective in $0.937 / Effective out $22.00
Token share
0.0%
Effective in $4.399 / Effective out $22.00
Token share
0.0%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Throughput
175tok/s
P50, best across providers
Latency
1.04s
P50, best provider
Reasoning effortLatency / throughput
AutoExacto benchmarks
GPQA DiamondTAU-BenchAmazon Bedrock (US)90.1%77.3%Claude Platform on AWS89.9%76.9%Google Vertex (US)90.0%76.2%Azure (US)90.3%75.5%Azure89.6%75.5%auto-routing90.1%75.0%Google Vertex (EU)90.0%74.9%Amazon Bedrock90.6%74.2%Amazon Bedrock (EU)89.8%74.9%Google Vertex89.8%73.2%
+1 more providers
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Uptime (3d)The model was reachable. Request routed to a provider.
100.00%
Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.
99.83%
Availability over the last 3 days
Last 72 hours
Availability 99.83%
3 Days Ago2 Days AgoYesterdayNow
Availability over the last 24 hours
OpenRouter Availability
99.94%
Without Routing
97.84%
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
Benchmark score summary for Anthropic: Claude Opus 5.5 (Artificial Analysis, Design Arena and Vals AI)
SourceBenchmarkScore
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) Intelligence Index48.2
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) HLE54.7%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) AA-LCR84.3%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) GDPval-AA54.5%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) CritPt27.7%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) SciCode59.3%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) Terminal-Bench 4.052.5%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) AA-Omniscience Accuracy63.9%
Artificial AnalysisClaude Opus 5.5 (Medium, Default Fallback) AA-Omniscience Non-Hallucination Rate33.1%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) Intelligence Index51.8
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) HLE55.6%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) AA-LCR82.7%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) GDPval-AA60.4%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) CritPt30.9%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) SciCode60.4%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) Terminal-Bench 4.056.6%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) AA-Omniscience Accuracy64.8%
Artificial AnalysisClaude Opus 5.5 (High, Default Fallback) AA-Omniscience Non-Hallucination Rate36.1%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) Intelligence Index34.8
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) HLE48.3%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) AA-LCR80.7%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) GDPval-AA36.9%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) CritPt17.7%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) SciCode58.6%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) Terminal-Bench 4.031.3%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) AA-Omniscience Accuracy63.3%
Artificial AnalysisClaude Opus 5.5 (Low, Default Fallback) AA-Omniscience Non-Hallucination Rate33.7%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) Intelligence Index59.1
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) HLE61.4%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) AA-LCR84.7%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) GDPval-AA68.3%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) CritPt31.7%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) SciCode66.9%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) Terminal-Bench 4.059.6%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) AA-Omniscience Accuracy65.5%
Artificial AnalysisClaude Opus 5.5 (Max, Default Fallback) AA-Omniscience Non-Hallucination Rate49.7%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) Intelligence Index57.4
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) HLE57.5%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) AA-LCR84.7%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) GDPval-AA66.8%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) CritPt31.7%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) SciCode65.0%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) Terminal-Bench 4.059.6%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) AA-Omniscience Accuracy65.5%
Artificial AnalysisClaude Opus 5.5 (Xhigh, Default Fallback) AA-Omniscience Non-Hallucination Rate35.0%
Design ArenaClaude Opus 5.5 Agents Arena Androidnative Elo1389
Design ArenaClaude Opus 5.5 Agents Arena Blender3d Elo1247
Design ArenaClaude Opus 5.5 Agents Arena Codetovideo Elo1336
Design ArenaClaude Opus 5.5 Agents Arena Full Stack Elo1390
Design ArenaClaude Opus 5.5 Agents Arena Mobile Apps Elo1294
Design ArenaClaude Opus 5.5 Agents Arena Pdf Elo1303
Design ArenaClaude Opus 5.5 Agents Arena Spreadsheet Elo1328
Design ArenaClaude Opus 5.5 Agents Arena Webapps Elo1321
Design ArenaClaude Opus 5.5 Models Arena 3D Elo1460
Design ArenaClaude Opus 5.5 Models Arena Asciiart Elo1357
Design ArenaClaude Opus 5.5 Models Arena Code Categories Elo1385
Design ArenaClaude Opus 5.5 Models Arena Data Visualization Elo1385
Design ArenaClaude Opus 5.5 Models Arena Game Development Elo1426
Design ArenaClaude Opus 5.5 Models Arena SVG Elo1321
Design ArenaClaude Opus 5.5 Models Arena UI Component Elo1375
Design ArenaClaude Opus 5.5 Models Arena Website Elo1326
Vals AIClaude Opus 5.5 Access to Justice Benchmark73.8%
Vals AIClaude Opus 5.5 BioMysteryBench79.3%
Vals AIClaude Opus 5.5 ClaimBench100.0%
Vals AIClaude Opus 5.5 Code Migration v1.181.0%
Vals AIClaude Opus 5.5 CyberBench v1.155.4%
Vals AIClaude Opus 5.5 CyberOffense77.1%
Vals AIClaude Opus 5.5 EMB v1.175.8%
Vals AIClaude Opus 5.5 Finance Agent v2.167.8%
Vals AIClaude Opus 5.5 Vals Finance Index65.3%
Vals AIClaude Opus 5.5 Financial Coaching Bench76.2%
Vals AIClaude Opus 5.5 HLAB-Vals-1.19.2%
Vals AIClaude Opus 5.5 Immigration72.1%
Vals AIClaude Opus 5.5 IOI95.1%
Vals AIClaude Opus 5.5 Vals Legal Index72.4%
Vals AIClaude Opus 5.5 Legal Research Bench50.5%
Vals AIClaude Opus 5.5 MedCode49.8%
Vals AIClaude Opus 5.5 MedScribe91.4%
Vals AIClaude Opus 5.5 MysteryMechanism49.5%
Vals AIClaude Opus 5.5 ProgramBench18.5%
Vals AIClaude Opus 5.5 ProofBench v1.1100.0%
Vals AIClaude Opus 5.5 Public Benefits Bench v1.170.6%
Vals AIClaude Opus 5.5 Robo-bench14.6%
Vals AIClaude Opus 5.5 Vals RSI Index37.3%
Vals AIClaude Opus 5.5 SAGE91.3%
Vals AIClaude Opus 5.5 Vals Science Index58.7%
Vals AIClaude Opus 5.5 SRE Bench33.6%
Vals AIClaude Opus 5.5 SWE-bench Multimodal70.2%
Vals AIClaude Opus 5.5 Tax Agent Bench44.5%
Vals AIClaude Opus 5.5 Terminal-Bench 4.065.2%
Vals AIClaude Opus 5.5 Terminal-Bench-Science47.1%
Vals AIClaude Opus 5.5 Vals Index70.1%
Vals AIClaude Opus 5.5 Vibe Code Bench 1-100 v1.130.9%
Vals AIClaude Opus 5.5 Vibe Code Bench v1.190.3%
Vals AIClaude Opus 5.5 Vals Web Search Index53.7%
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
1.
Claude Code
Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.
13.9Ttokens
2.
Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
1.28Ttokens
3.
pi
There are many coding agents, but this one is yours.
894Btokens
4.
omp
new
391Btokens
5.
Codex
A coding agent that helps you build and ship with AI
159Btokens
Rest of OpenRouter
22.8B
Top 5 Apps
18.6B
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
Anthropic: Claude Opus versions
Available through EU in-region routing
Anthropic: Claude Opus 5.5 is available through EU in-region routing. Requests sent to eu.openrouter.ai are decrypted in the EU and routed only to shared provider endpoints in Europe (the EU/EEA, the UK, and Switzerland).
Providers in EuropeAmazon Bedrock, Google VertexPrices in Europe
Amazon Bedrock: $4.40/M tokens input, $22/M tokens output
Google Vertex: $4.40/M tokens input, $22/M tokens output
Base URLhttps://eu.openrouter.ai/api/v1PlanRequires Business or Enterprise. Self-serve on Business: no contract, no monthly minimum. OpenRouter Business for EU in-region routing
Fails closed: if no shared provider endpoint in Europe is available, the request errors instead of leaving the region. Requests served by your own BYOK key go to your deployment’s region, which you control. In-region routing docs
Documentation
Guides, cookbooks, blog posts, and tools for building with Anthropic: Claude Opus 5.5 on OpenRouter.
Claude 5.5 Migration GuideCookbook
Best OCR Model for Scanned Documents, MeasuredBlog post
Claude Opus vs Sonnet vs Haiku: Which Should You Use?Blog post
How Much Does Claude Code Cost? Real Session Costs by ModelBlog post
Hit Your Claude Code Usage Limit? How to Keep CodingBlog post
Claude 529 Overloaded: What It Means and How to Fix ItBlog post
Frequently asked questions
Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code review and bug finding, financial and scientific analysis, and reading dense charts, diagrams, and screenshots, and it is more careful than its predecessor about only stating figures and...
The Claude Opus 5.5 documentation on OpenRouter includes the Claude 5.5 Migration Guide cookbook, the Best OCR Model for Scanned Documents, Measured blog post, the Claude Opus vs Sonnet vs Haiku: Which Should You Use? blog post, the How Much Does Claude Code Cost? Real Session Costs by Model blog post, the Hit Your Claude Code Usage Limit? How to Keep Coding blog post, and the Claude 529 Overloaded: What It Means and How to Fix It blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.
Claude Opus 5.5 costs $4.00/M input tokens and $20.00/M output tokens, with separate rates for Cache Read at $0.20/M tokens, Cache Write at $5.00/M tokens, and Cache Write (1h) at $8.00/M tokens.
Claude Opus 5.5 has a 1,000,000 token context window. It supports up to 128,000 completion tokens.
Yes. Claude Opus 5.5 accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
Claude Opus 5.5 accepts text, images, and files such as PDFs as input and returns text.
Claude Opus 5.5 is served by 5 providers on OpenRouter: Amazon Bedrock, Azure, Google Vertex, Claude Platform on AWS, and Anthropic. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
Claude Opus 5.5 was released on September 22, 2026.