Day 0 GPT-5.6 release

[mateo-berri]

GPT-5.6 is now live on LiteLLM

Call it through the LiteLLM AI Gateway using the same OpenAI-compatible request you already use, with spend tracking, rate limits, fallbacks, and logging in one place. No SDK changes, no per-provider rewrites.

What's new:

• One family, three sizes: GPT-5.6 ships as gpt-5.6-sol (the frontier tier for complex professional work), gpt-5.6-terra (the balanced mid tier), and gpt-5.6-luna (the fastest, cheapest tier). The bare gpt-5.6 alias routes to Sol.
• Frontier reasoning, fewer tokens: Sol reaches frontier performance with lower output-token counts, adds a max reasoning-effort setting for the hardest tasks, and introduces a pro reasoning mode when quality matters more than latency. Prompting that works on GPT-5.5 carries straight over.
• 1.05M context, 128K output: the full family shares a 1,050,000-token context window and 128,000-token max output, with text and image input.
• Pricing (standard, per MTok): Sol is $5 in / $30 out, Terra is $2.50 in / $15 out, Luna is $1 in / $6 out. Cache reads are $0.50 / $0.25 / $0.10. Cache writes are new for this family, billed at 1.25x the uncached input rate: $6.25 / $3.125 / $1.25. Prompts over 272K input tokens are priced at 2x input and 1.5x output for the whole request; LiteLLM tracks all of it automatically, including Flex, Batch, and Priority tiers.

Get started, no redeploy needed:

GPT-5.6 is already in the LiteLLM model cost map. Reload your pricing in the UI under Models + Endpoints → Price Data → Reload Price Data (or POST /reload/model_cost_map as an admin). That enables the GPT-5.6 family with cost tracking and the full effort ladder.

model_list:
  - model_name: gpt-5.6
    litellm_params:
      model: openai/gpt-5.6
      api_key: os.environ/OPENAI_API_KEY
  - model_name: gpt-5.6-terra
    litellm_params:
      model: openai/gpt-5.6-terra
      api_key: os.environ/OPENAI_API_KEY
  - model_name: gpt-5.6-luna
    litellm_params:
      model: openai/gpt-5.6-luna
      api_key: os.environ/OPENAI_API_KEY

Running with LITELLM_LOCAL_MODEL_COST_MAP=true? Pull ghcr.io/berriai/litellm:v1.93.0-stable or later for the bundled metadata.

Read the full guide → Day 0 Support: GPT-5.6

Mateo Wang
LiteLLM AI Team

---

Please use this space to voice thoughts and opinions, or any issues you find

[Kefan-pauline]

Hi, thanks for adding the support.

azure/gpt-5.6 isn't available in the model_prices_and_context_window.json in main (only available in staging branch), so streaming isn't working correctly. Please promote the change timely, thanks!

[krrish-berri-2]

@mateo-berri

[mateo-berri]

Hi @Kefan-pauline, this has been addressed in 10d5804. Thanks for the heads up