GPT-5 mini vs Claude Haiku 4.5
Choose based on your ecosystem. GPT-5 mini for OpenAI integrations; Haiku 4.5 for Anthropic stack or 200K context needs.
GPT-5 mini ($0.25/M input) and Claude Haiku 4.5 ($0.80/M input) are both small-model cost leaders. GPT-5 mini is significantly cheaper. Haiku 4.5 has 200K context vs 128K, and performs better on non-English languages. At the same token volume, GPT-5 mini costs about 3× less.
Side-by-side pricing
| GPT-5 mini | Claude Haiku 4.5 | |
|---|---|---|
| Provider | OpenAI | Anthropic |
| Input price/1M | $0.25 | $1.00 |
| Output price/1M | $2.00 | $5.00 |
| Cache read/1M | $0.025 | $0.10 |
| Context window | 400K | 200K |
| Vision | ✓ | ✓ |
| Tool use | ✓ | ✓ |
| Prompt caching | ✓ | ✓ |
| Speed tier | Fast | Fast |
Prices verified 2026-08-20. See changelog for history.
Winner by task
Calculate costs for each model
Frequently Asked Questions
Is GPT-5 mini better than Claude Haiku 4.5?
GPT-5 mini is 3× cheaper on input tokens. Claude Haiku 4.5 has a larger context window (200K vs 128K) and better multilingual performance. For English-only classification and extraction at high volume: GPT-5 mini. For anything needing large context or non-English: Haiku 4.5.
Prices verified 2026-08-20. See methodology for calculation details.