Claude Haiku 4.5 vs Gemini 3.1 Flash-Lite: which API costs less?
Averaged across three realistic workloads, Gemini 3.1 Flash-Lite is about 49% cheaper than Claude Haiku 4.5. Below: raw rates and side-by-side monthly costs, computed from prices verified 2026-07-05 (we track what you are billed today, promos included).
Rates per 1M tokens
| Model | Input | Output | Cached input |
|---|---|---|---|
| Claude Haiku 4.5Anthropic | $1 | $5 | $0.100 |
| Gemini 3.1 Flash-LiteGoogle | $0.250 | $1.50 | — |
Monthly cost, side by side
| Workload | Claude Haiku 4.5 | Gemini 3.1 Flash-Lite | Difference |
|---|---|---|---|
Customer-support chatbot 2M input + 0.5M output tokens/month, 60% of input served from prompt cache | $3.42 | $1.25 | Gemini 3.1 Flash-Lite −63% |
RAG / document Q&A 20M input + 1M output tokens/month, 70% cached (shared context) | $12.40 | $6.50 | Gemini 3.1 Flash-Lite −48% |
Agentic coding assistant 50M input + 5M output tokens/month, 80% cached (long sessions) | $39.00 | $20.00 | Gemini 3.1 Flash-Lite −49% |
Your traffic looks different? Run your own numbers in the calculator.
FAQ
Which is cheaper: Claude Haiku 4.5 or Gemini 3.1 Flash-Lite?
Across three realistic workloads (chatbot, RAG, agent), Gemini 3.1 Flash-Lite works out about 49% cheaper than Claude Haiku 4.5 on average. Raw rates: Claude Haiku 4.5 $1/$5 per 1M tokens vs Gemini 3.1 Flash-Lite $0.250/$1.50. Verified 2026-07-05.
Does prompt caching change the comparison?
Yes — cached input costs $0.100 on Claude Haiku 4.5 and the full input rate on Gemini 3.1 Flash-Lite (no cache discount listed). Our scenario numbers already include realistic cache hit rates.
Is cheaper always better?
No. Price is one axis — model capability, context window and latency differ. This page answers only the cost question; benchmark quality on your own tasks before switching.