Gemini 2.5 Flash vs Claude Haiku 4.5
Gemini 2.5 Flash is significantly cheaper. Haiku 4.5 wins on multilingual tasks and Anthropic ecosystem.
Gemini 2.5 Flash ($0.075/M) vs Claude Haiku 4.5 ($0.80/M) — Flash is 10× cheaper on input. Both are fast, capable small models. Flash has 1M context vs Haiku's 200K. Haiku wins on non-Latin multilingual tasks and for teams already using Anthropic's ecosystem. At pure cost, Flash is the clear winner.
Side-by-side pricing
| Gemini 2.5 Flash | Claude Haiku 4.5 | |
|---|---|---|
| Provider | Anthropic | |
| Input price/1M | $0.30 | $1.00 |
| Output price/1M | $2.50 | $5.00 |
| Cache read/1M | $0.075 | $0.10 |
| Context window | 1M | 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 Gemini Flash 10× better value than Haiku?
On pure cost per token: yes. On task quality: Flash and Haiku are comparable for English tasks, with Haiku ahead on non-Latin scripts. If you're building on the Anthropic API and want to minimize provider switches, stick with Haiku. If you're cost-optimizing and English-only: Flash is compelling.
Prices verified 2026-08-20. See methodology for calculation details.