Skip to main content

Anthropic Claude Setup

Anthropic is available in AxonFlow Community and Enterprise. It is a strong option for long-context workflows, reasoning-heavy agent tasks, and teams that prefer Claude as the primary model family.

Runtime Defaults​

AxonFlow's Anthropic provider defaults to:

  • Provider name: anthropic
  • Default model: claude-sonnet-4-20250514
  • Default base URL: https://api.anthropic.com

Current model aliases in the runtime map older Claude constants onto the current-generation Claude 4 family, so it is still better to set the exact model you want in configuration.

Environment Variables​

export ANTHROPIC_API_KEY=sk-ant-your-api-key
export ANTHROPIC_MODEL=claude-sonnet-4-20250514
export ANTHROPIC_TIMEOUT_SECONDS=120    # Request timeout in seconds (default: 120)
export ANTHROPIC_ENDPOINT=https://api.anthropic.com  # Custom endpoint for proxies

YAML Configuration​

version: "1.0"

llm_providers:
  anthropic:
    enabled: true
    credentials:
      api_key: ${ANTHROPIC_API_KEY}
    config:
      model: ${ANTHROPIC_MODEL:-claude-sonnet-4-20250514}

Good Fits​

  • Research and reasoning workflows
  • Long-context enterprise copilots
  • Teams that want a safety-focused Claude default in community mode
  • Knowledge assistants, analysis pipelines, and review workflows that need Claude capability with production governance

Proxy Mode​

import { AxonFlow } from '@axonflow/sdk';

const axonflow = new AxonFlow({
  endpoint: 'http://localhost:8080',
  clientId: process.env.AXONFLOW_CLIENT_ID,
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET,
});

const response = await axonflow.proxyLLMCall({
  userToken: 'user-123',
  query: 'Explain the security tradeoffs of this architecture.',
  requestType: 'chat',
  context: {
    provider: 'anthropic',
    model: 'claude-sonnet-4-20250514',
  },
});

console.log(response.data);

Gateway Mode​

import { AxonFlow } from '@axonflow/sdk';
import Anthropic from '@anthropic-ai/sdk';

const prompt = 'Explain the security tradeoffs of this architecture.';

const axonflow = new AxonFlow({
  endpoint: 'http://localhost:8080',
  clientId: process.env.AXONFLOW_CLIENT_ID,
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET,
});

const anthropic = new Anthropic({ apiKey: process.env.ANTHROPIC_API_KEY });

const ctx = await axonflow.getPolicyApprovedContext({
  userToken: 'user-123',
  query: prompt,
});

if (!ctx.approved) {
  throw new Error(`Blocked: ${ctx.blockReason}`);
}

const approvedPrompt =
  typeof ctx.approvedData.query === 'string' ? String(ctx.approvedData.query) : prompt;

const message = await anthropic.messages.create({
  model: 'claude-sonnet-4-20250514',
  max_tokens: 1024,
  messages: [{ role: 'user', content: approvedPrompt }],
});

const firstBlock = message.content.find(block => block.type === 'text');
const output = firstBlock?.text ?? '';

await axonflow.auditLLMCall({
  contextId: ctx.contextId,
  responseSummary: output.slice(0, 200),
  provider: 'anthropic',
  model: 'claude-sonnet-4-20250514',
  tokenUsage: {
    promptTokens: message.usage.input_tokens,
    completionTokens: message.usage.output_tokens,
    totalTokens: message.usage.input_tokens + message.usage.output_tokens,
  },
  latencyMs: 250,
});

Notes for Production Teams​

  • The Anthropic provider supports streaming in the runtime.
  • For request-level routing, context.provider = "anthropic" is a preference unless strict pinning is enabled for that request.
  • If your organization standardizes on Claude but needs runtime routing, combine Anthropic with a secondary provider and configure failover or weighted routing at the platform level.

Troubleshooting​

IssueCauseFix
401 Authentication failedInvalid API keyVerify ANTHROPIC_API_KEY
429 Rate limitedConcurrent request limitReduce concurrency or contact Anthropic
Overloaded errorAnthropic capacity issuesRetry with backoff

See Also​

Provider Rollout Checklist​

Before this provider path becomes production traffic, connect it to the runtime operating model:

Last updated on