// head_to_head
Gemini 3.5 Flash Lite vs Qwen2.5 VL 72B Instruct
Published LiveBench scores across all seven categories, live list pricing, context windows, and the measured cost of a point of capability — for both models, side by side.
Gemini 3.5 Flash Lite and Qwen2.5 VL 72B Instruct are priced within ~10% of each other.
Qwen2.5 VL 72B Instruct does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.
Gemini 3.5 Flash Lite
- Blended / 1M
- $0.850
- Context
- 1.0M
- Released
- Jul 21, 2026
- Overall score
- 63.9
qwen
Qwen2.5 VL 72B Instruct
- Blended / 1M
- $0.850
- Context
- 128K
- Released
- Feb 1, 2025
- Overall score
- Not evaluated
Specs and pricing
| Metric | Gemini 3.5 Flash Lite | Qwen2.5 VL 72B Instruct |
|---|---|---|
| LiveBench overall Mean of the seven LiveBench category scores, 0–100. Higher is better. | 63.9 | — |
| Cost per point Measured benchmark spend divided by overall score — dollars per point of capability. | $0.0379 | — |
| Blended price / 1M 3:1 input:output mix, the usual shape of production traffic. | $0.850 | $0.850 |
| Input price / 1M | $0.300win | $0.800 |
| Output price / 1M | $2.50 | $1.00win |
| Cached input / 1M Price of an input token served from the prompt cache, where the provider publishes one. | $0.030win | $0.400 |
| Context window | 1.0Mwin | 128K |
| Max output tokens | 66K | 115Kwin |
Benchmarks by category
An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Gemini 3.5 Flash Lite on top, Qwen2.5 VL 72B Instruct below, both out of 100.
What each one costs to run
Per-token prices are hard to feel. These are monthly list costs for both models across five workload shapes, using each provider's published cached-input rate where there is one.
| Workload | Gemini 3.5 Flash Lite | Qwen2.5 VL 72B Instruct |
|---|---|---|
| Support chatbot 1.2K in / 400 out × 200K requests | $252.56/mo | $243.20/mo |
| RAG assistant 8K in / 600 out × 100K requests | $282.00/mo | $540.00/mo |
| Coding agent 40K in / 4K out × 20K requests | $288.80/mo | $496.00/mo |
| Document extraction 20K in / 1.5K out × 50K requests | $474.00/mo | $855.00/mo |
| Bulk classification 500 in / 20 out × 5M requests | $865.00/mo | $1,900/mo |
Which should you pick?
You need to fit large documents in one call
Gemini 3.5 Flash Lite
Wider context window — 1.0M against 128K.
Gemini 3.5 Flash Lite vs Qwen2.5 VL 72B Instruct FAQ
Which is better, Gemini 3.5 Flash Lite or Qwen2.5 VL 72B Instruct?
Gemini 3.5 Flash Lite and Qwen2.5 VL 72B Instruct are priced within ~10% of each other. Qwen2.5 VL 72B Instruct does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.
Is Gemini 3.5 Flash Lite cheaper than Qwen2.5 VL 72B Instruct?
They cost about the same. Both land near $0.850 per million tokens on a 3:1 input:output blend, so price is unlikely to be the deciding factor between them.
Does Gemini 3.5 Flash Lite or Qwen2.5 VL 72B Instruct have a bigger context window?
Gemini 3.5 Flash Lite has the larger context window: 1.0M for Gemini 3.5 Flash Lite against 128K for Qwen2.5 VL 72B Instruct. Note that a window you can fill is not a window you should fill — retrieval quality usually degrades well before the limit, and you pay for every token you put in it.
Do Gemini 3.5 Flash Lite and Qwen2.5 VL 72B Instruct support prompt caching?
Both publish a cached-input rate: $0.030 per million for Gemini 3.5 Flash Lite and $0.400 for Qwen2.5 VL 72B Instruct, against full input rates of $0.300 and $0.800. On a workload with a long stable prefix — a system prompt, a tool schema, a retrieved corpus — that changes the economics more than the headline price does.
Related comparisons
How these numbers are produced
- Price — provider list price from OpenRouter, refreshed every 15 minutes. “Blended” is a 3:1 input:output mix.
- Scores — LiveBench release 2026-06-25, using their own category map. Each model shows its strongest published run. A blank means “not evaluated”, never “bad”.
- Cost per point — the measured dollars LiveBench spent on the run, divided by the score it earned.
- “Win” — awarded only past a threshold: one full point on a benchmark score, 10% on a price, 25% on a context window. Anything tighter reports as a tie, because effort settings alone move a LiveBench score by more than that.
Published benchmarks rank models on someone else's tasks. Before committing, see LLM & agent evaluation for building an eval on your own.