### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### The Feature

Add cost-map entries for **`gemini-3.1-flash-lite-image`** (Google's *Nano Banana 2 Lite*) — including the `gemini/` and `vertex_ai/` variants — to `model_prices_and_context_window.json` (and the bundled backup).

The sibling models are already mapped (`gemini-3.1-flash-image`, `gemini-3.1-flash-lite`), but the **lite image** model is missing.

Proposed values (Google list price):

| field | value | note | |---|---|---| | `input_cost_per_token` | `2.5e-07` | $0.25/1M | | `output_cost_per_token` | `1.5e-06` | $1.50/1M (text/thinking) | | `output_cost_per_image_token` | `3e-05` | $30/1M (image output) | | `output_cost_per_image` | `0.0336` | 1120 tokens × 3e-5 (1K image) | | `input_cost_per_image` | `0.00028` | 1120 tokens × 2.5e-7 | | `mode` | `image_generation` | |

Source: https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models — mirrors the existing `gemini-3.1-flash-image` entry, with costs at the Lite tier (exactly half the non-lite image model).

### Motivation, pitch

Because this model isn't mapped, `litellm.completion_cost(m…