Day 0 support for Gemini 3.8 Flash on LiteLLM!

[misbahsy]

Gemini 3.8 Flash is now live on LiteLLM

What's new:

Reasoning and coding close to larger frontier models - Google reports gains on long-horizon software engineering, stronger multi-step reasoning in finance and legal domains, and 54.9% on HLE-Verified.

Half price through December 31, 2026 - $0.75 / MTok input and $3.75 / MTok output, going to $1.50 / $7.50 on January 1, 2027. Cache reads are $0.075 / MTok, batch and flex run at half the standard rate, and LiteLLM tracks cost at the promotional rate.

1M context, 65K output - a 1,048,576-token context window and 65,536-token max output, with full multimodal input (text, image, audio, video), function calling, prompt caching from 4,096 tokens, and web search.

Get started, no redeploy needed:

Gemini 3.8 Flash is in the LiteLLM model cost map as of PR #39340. Reload your pricing in the UI under Models + Endpoints -> Price Data -> Reload Price Data (or POST /reload/model_cost_map as an admin) and add the following model entry.

model_list:
  - model_name: gemini-3.8-flash
    litellm_params:
      model: gemini/gemini-3.8-flash
      api_key: os.environ/GEMINI_API_KEY

Read the full guide → Day 0 Support: Gemini 3.8 Flash

Misbah Syed
DevRel @ LiteLLM