Smart Router Configuration & Cost Failover: Claude 3.5 Haiku (anthropic/claude-3.5-haiku) to Mistral Codestral 2501 (mistralai/codestral-2501)

Production LLM API routing configuration, TTFT latency SLA (<400ms), and contribution margin failover benchmark generated by TokenMargin.dev.

1. Pricing & Latency Benchmark Summary

• Primary Model (Claude 3.5 Haiku — anthropic/claude-3.5-haiku): $0.80 / 1M input tokens, $4.00 / 1M output tokens | TTFT: 340ms | Throughput: 85 tok/s
• Fallback Model (Mistral Codestral 2501 — mistralai/codestral-2501): $0.30 / 1M input tokens, $0.90 / 1M output tokens | TTFT: 340ms | Throughput: 85 tok/s

2. Zero-Dependency TypeScript / Node.js Smart Router Configuration

export const SMART_ROUTER_CONFIG = {
  slaMaxTtftMs: 400,
  primary: { model: "anthropic/claude-3.5-haiku", inputPer1M: 0.80, outputPer1M: 4.00 },
  fallback: { model: "mistralai/codestral-2501", inputPer1M: 0.30, outputPer1M: 0.90 },
  retryOnStatus: [429, 500, 502, 503, 529]
};

3. LiteLLM / OpenRouter YAML Router Configuration (config.yaml)

model_list:
  - model_name: production-primary
    litellm_params:
      model: openrouter/anthropic/claude-3.5-haiku
      timeout: 15
  - model_name: cost-fallback
    litellm_params:
      model: openrouter/mistralai/codestral-2501
router_settings:
  routing_strategy: latency-based-routing
  fallbacks: [{"production-primary": ["cost-fallback"]}]