Use one CorvusLLM key with supported Claude model rows for coding agents, research, writing, and automation through an OpenAI-compatible proxy route.
Independent service. Not affiliated with OpenAI, Anthropic, Google, or Z.AI.
This page is the Claude-family buying and selection page. Use it to compare current CorvusLLM Claude rows, public slugs, prepaid prices, cache pricing context, and the safer setup path before sending real project context.
Model pages are for selection. Exact technical setup, live availability, and private account state belong to the linked sources.
| Question | Use this source | Why it matters |
|---|---|---|
| Which slug should I send? | data/models.json and model docs | They expose customer-facing slugs and avoid hidden route, backup, or provider-account names. |
| What will it cost? | cost calculator and pricing tracker | Input, output, cache read, and cache write can change the final billed amount. |
| How do I configure my tool? | Claude Code setup | Client setup depends on whether the tool expects OpenAI-compatible or provider-native fields. |
This matrix turns the model list into a decision table for searchers comparing quality, cost, and fit before they buy.
| Model row | Best fit | Check first | Input | Output | Cache read |
|---|---|---|---|---|---|
| Claude Haiku 4.5claude-haiku-4-5 | Fast lightweight summarization, extraction, and automation. | Avoid when complex reasoning, large code changes, or premium writing quality are required. | $0.350/1M | $1.75/1M | $0.035/1M |
| Claude Opus 4.5claude-opus-4-5 | Highest-quality coding, research, and multi-step reasoning work. | Avoid for very cheap high-volume extraction unless quality justifies the spend. | $1.75/1M | $8.75/1M | $0.175/1M |
| Claude Opus 4.6claude-opus-4-6 | Highest-quality coding, research, and multi-step reasoning work. | Avoid for very cheap high-volume extraction unless quality justifies the spend. | $1.75/1M | $8.75/1M | $0.175/1M |
| Claude Opus 4.7claude-opus-4-7 | Highest-quality coding, research, and multi-step reasoning work. | Avoid for very cheap high-volume extraction unless quality justifies the spend. | $1.75/1M | $8.75/1M | $0.175/1M |
| Claude Opus 4.8claude-opus-4-8 | Highest-quality coding, research, and multi-step reasoning work. | Avoid for very cheap high-volume extraction unless quality justifies the spend. | $1.75/1M | $8.75/1M | $0.175/1M |
| Claude Fable 5claude-fable-5 | General CorvusLLM model-family workloads. | Run a small test before using it for production traffic. | $5.00/1M | $20.00/1M | $3.00/1M |
| Claude Sonnet 4.5claude-sonnet-4-5 | Balanced agent, coding, writing, and analysis workloads. | Use Opus instead when the task needs maximum depth; use Haiku for the cheapest short jobs. | $1.05/1M | $5.25/1M | $0.105/1M |
| Claude Sonnet 4.6claude-sonnet-4-6 | Balanced agent, coding, writing, and analysis workloads. | Use Opus instead when the task needs maximum depth; use Haiku for the cheapest short jobs. | $1.05/1M | $5.25/1M | $0.105/1M |
| Claude Sonnet 5claude-sonnet-5 | Balanced agent, coding, writing, and analysis workloads. | Use Opus instead when the task needs maximum depth; use Haiku for the cheapest short jobs. | $1.05/1M | $5.25/1M | $0.105/1M |
These are the stable public setup values for Claude workloads. Client-specific screens are explained in the linked setup docs.
| Configuration field | Value | When to use it |
|---|---|---|
| OpenAI-compatible base URL | https://base.corvusllm.com/v1 |
Use this for OpenAI SDKs, curl, fetch, Open WebUI, n8n, ChatBox, and tools with a custom OpenAI-compatible endpoint field. |
| Anthropic-native base URL | https://base.corvusllm.com/anthropic |
Use this only for Claude Code style clients that expect Anthropic-compatible request behavior. |
| Authorization header | Bearer YOUR_CORVUSLLM_KEY |
Paste the delivered prepaid key. Do not expose it in browser-side production code or public repositories. |
| Model parameter | Use the public slug from this page |
Use customer-facing slugs only. Do not send private route names, upstream account names, or backup identifiers. |
| First test | Send one low-risk prompt |
Confirm output quality, response status, and billed usage before sending real project context. |
| Pricing check | /llm-api-cost-calculator |
Estimate input, output, cache read, and cache write before bulk traffic. |
These rows mirror the public model catalog and pricing tracker. Always confirm live availability before moving production traffic.
Fast Claude row for lightweight automation, summarization, and extraction.
model: claude-haiku-4-5Highest-capability Claude row for deep coding, research, and multi-step reasoning.
model: claude-opus-4-5Highest-capability Claude row for deep coding, research, and multi-step reasoning.
model: claude-opus-4-6Highest-capability Claude row for deep coding, research, and multi-step reasoning.
model: claude-opus-4-7Highest-capability Claude row for deep coding, research, and multi-step reasoning.
model: claude-opus-4-8Claude Fable 5 is part of the current CorvusLLM public model catalog.
model: claude-fable-5Balanced Claude row for coding agents, writing, analysis, and day-to-day assistant work.
model: claude-sonnet-4-5Balanced Claude row for coding agents, writing, analysis, and day-to-day assistant work.
model: claude-sonnet-4-6Balanced Claude row for coding agents, writing, analysis, and day-to-day assistant work.
model: claude-sonnet-5CorvusLLM deducts usage from prepaid balance using its listed input, output, cache-read, and cache-write rates where applicable. Official prices are shown as comparison references, not invoices from the provider.
Do not send sensitive or regulated data through shared API proxies. Use direct providers or enterprise-reviewed infrastructure for regulated workloads.
Clear limits reduce mismatched traffic and make the page more useful than a generic model catalog.
Lowest-cost high-volume extraction where a cheaper public catalog row gives acceptable quality.
Official Anthropic procurement, enterprise support, or direct Anthropic account management.
Sensitive, regulated, or confidential data that should not pass through a shared proxy.
Start with the setup guide that matches your tool, then switch only the model parameter when changing rows in this family.
Use the calculator and source-linked tracker together before moving real workloads onto a third-party proxy route.
These pages explain service limits, support paths, and operational evidence without implying official provider affiliation.
Short answers for buyers, developers, and AI assistants that need the canonical setup path.
Use https://base.corvusllm.com/v1 for OpenAI-compatible clients and https://base.corvusllm.com/anthropic for Claude Code style Anthropic-native clients.
No. You use a prepaid CorvusLLM key, but CorvusLLM is an independent proxy and not an official Anthropic account, partner, subscription, or support contract.
Start with Sonnet for balanced work, Opus for maximum quality, and Haiku for lightweight tasks.
Read Claude Code setup, confirm the exact public slug, run one low-risk request, check billed usage, and avoid sensitive or regulated data on shared proxy infrastructure.
Use a low-risk prompt, confirm the exact model slug, then compare output quality and billed usage before larger workloads.
Move from model selection to exact slugs, cost estimates, billing behavior, and service limits without relying on old screenshots.