AI API Cost Optimization for Developers

AI API Cost Optimization for Developers

AI API costs can spiral quickly. Understanding token pricing, caching strategies, and model selection helps you build AI features sustainably.

Understanding Token Pricing

LLM APIs charge per token (roughly 4 characters or 0.75 words). Most differentiate between input and output tokens.

2026 Approximate Pricing (per 1M tokens)

| Model | Input | Output |

|-------|-------|--------|

| GPT-4o | $2.50 | $10.00 |

| Claude Sonnet | $3.00 | $15.00 |

| Claude Haiku | $0.25 | $1.25 |

| Gemini Pro | $1.25 | $5.00 |

| GPT-4o mini | $0.15 | $0.60 |

Cost Reduction Strategies

1. Choose the Right Model

Not every task needs the most powerful model. Route by complexity:

Simple classification/extraction → GPT-4o mini or Haiku ($0.15-0.25/1M input)

Standard generation → GPT-4o or Sonnet ($2.50-3.00/1M input)

Complex reasoning → Claude Opus or GPT-4o ($5-15/1M input)

2. Minimize Input Tokens

Trim context — don't send your entire codebase. Send only relevant files. Compress prompts — remove redundant instructions. "Respond in JSON" is shorter than "Please format your response as a JSON object with the following structure..." Use system prompts wisely — they're cached in some APIs (Claude's prompt caching), so put stable instructions there.

3. Cache Responses

Cache identical or similar requests:

const cacheKey = crypto.createHash("md5")
  .update(JSON.stringify({ model, messages }))
  .digest("hex");

const cached = await redis.get(cacheKey);
if (cached) return JSON.parse(cached);

const response = await openai.chat.completions.create({ model, messages });
await redis.set(cacheKey, JSON.stringify(response), "EX", 3600);

4. Use Streaming

Streaming doesn't reduce cost, but it improves perceived performance. Users see the first token in ~200ms instead of waiting 5-10 seconds for the full response.

5. Set Max Tokens

Always set max_tokens to prevent runaway responses:

const response = await openai.chat.completions.create({
  model: "gpt-4o",
  messages,
  max_tokens: 500,  // Cap output length
});

6. Batch Processing

Use batch APIs for non-real-time workloads. OpenAI's batch API is 50% cheaper.

Monitoring Costs

Track token usage per request, user, and feature:

const usage = response.usage;
console.log({
  inputTokens: usage.prompt_tokens,
  outputTokens: usage.completion_tokens,
  estimatedCost: (usage.prompt_tokens * 2.5 + usage.completion_tokens * 10) / 1_000_000,
});

Try It

Use our AI Cost Calculator to compare pricing across models and estimate costs for your expected volume.

Conclusion

AI costs are manageable with the right strategy: use cheaper models for simple tasks, cache responses, minimize context, and monitor usage. The difference between a naive and optimized implementation can be 10x.