budget
Claude Haiku 4 5 vs Gemini 2.5 Flash
Side-by-side pricing, context window, features, and per-use-case verdict. Data sourced from LiteLLM, updated regularly.
Specifications
| Feature | Claude Haiku 4 5 | Gemini 2.5 Flash |
|---|---|---|
| Provider | anthropic | vertex ai-language-models |
| Model ID | claude-haiku-4-5 | gemini-2.5-flash |
| Input / 1M tokens | $1.00 | $0.30 (70% cheaper) |
| Output / 1M tokens | $5.00 | $2.50 (50% cheaper) |
| Cached input / 1M | $0.10 | $0.03 |
| Context window | 200kk tokens | 1049kk tokens |
| Max output | 64kk tokens | 66kk tokens |
| Vision | ✅ | ✅ |
| Function calling | ✅ | ✅ |
| Prompt caching | ✅ | ✅ |
| Web search | ⬜ | ✅ |
| JSON schema | ✅ | ✅ |
Monthly cost comparison
No cache discount applied — pure token costs at three usage scales.
| Scale | Claude Haiku 4 5 | Gemini 2.5 Flash | Winner |
|---|---|---|---|
| Small 1M in / 0.2M out | $2.00 | $0.80 | Gemini 2.5 Flash saves 60% |
| Medium 10M in / 2M out | $20.00 | $8.00 | Gemini 2.5 Flash saves 60% |
| Large 100M in / 20M out | $200 | $80.00 | Gemini 2.5 Flash saves 60% |
Category winners
Cheapest input
Gemini 2.5 Flash
$0.30/M
· 70% less than the other
Cheapest output
Gemini 2.5 Flash
$2.50/M
· 50% less than the other
Larger context
Gemini 2.5 Flash
1049kk tokens · 424% larger window
More features
Gemini 2.5 Flash
5 / 5 capabilities
Better for…
High-volume RAG & repeated context
$0.03/M cached input — prompt caching makes repeated context retrieval significantly cheaper
Gemini 2.5 Flash
Short interactions & high-throughput APIs
$2.50/M output — 50% cheaper on output at scale
Gemini 2.5 Flash
Long-document processing & large codebases
1049k context window vs 200k — 424% larger context
Gemini 2.5 Flash
Agentic workflows & tool use
Supports 5 of 5 tracked capabilities (vision, function calling, caching, web search, JSON schema)
Gemini 2.5 Flash
Deep-dive on each model
Pricing data sourced from LiteLLM and refreshed regularly. Always verify with each provider's official pricing page before making business decisions.