---
title: "1.105.0rc3 - Decisions API"
url: "/release_notes/v1.105.0rc3/v1-105-0-rc-3"
canonical_url: "https://docs.litellm.ai/release_notes/v1.105.0rc3/v1-105-0-rc-3"
type: "release-notes"
last_updated: "2026-10-08"
related:
  - "/release_notes/v1.105.0rc1/v1-105-0-rc-1"
  - "/release_notes/v1.104.2/v1-104-2"
---
# 1.105.0rc3 - Decisions API

> Index of all LiteLLM docs: https://docs.litellm.ai/llms.txt

## Deploy this version

**Docker**

```bash
docker run \
-e LITELLM_MASTER_KEY=sk-<paste-a-long-random-key> \
-e DATABASE_URL=postgresql://<user>:<password>@<host>:5432/<dbname> \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
docker.litellm.ai/berriai/litellm:1.105.0-rc.3
```

**Pip**

```bash
pip install litellm==1.105.0rc3
```

The published GitHub tag is `v1.105.0-rc.3`. These notes compare it with [`v1.105.0-rc.2`](https://github.com/BerriAI/litellm/releases/tag/v1.105.0-rc.2), the previous release candidate on `rc/1.105.0`. It adds the Decisions API routes and providers. There are no new database migrations, and the Decisions routes are new on this line. The `v1.105.0-rc.3` tag points at [`c09b4d5`](https://github.com/BerriAI/litellm/commit/c09b4d5a09e6f5c09080c0762fb200d91572f331)

## Decisions API

The proxy now serves decision models, which answer typed questions (yes/no predicates, choices and scores) about an input. Requests go through the usual keys, budgets, spend tracking and guardrails. See the [Decisions docs](../../docs/decisions)

- **Two request formats**
    - `POST /v1/decisions` and `POST /decisions` take OpenAI's Decisions API format: `input` plus a `questions` list, answered as an `answers` list with OpenAI-style `usage`
    - `POST /v1/systemone` and `POST /systemone` take the System One format: `state` plus a named `questions` map, answered as an `answers` map keyed by question name
    - Both routes go through one shared format inside the proxy, so any Decisions model can be called on either route
- **Providers**: `openai/` (`gpt-6-luna`), `perplexity/`, `typesafe/`, `openrouter/`, `cloudflare/clef` and `cloudflare/clef-flash`, and self-hosted `strands_decider/` (requires `api_base`)
- **OpenAI**: images, multi-message input, `safety_identifier`, names and refusals pass through on `/v1/decisions`. The provider reads `litellm.api_key`, `litellm.openai_key`, `litellm.api_base`, `OPENAI_BASE_URL` and `OPENAI_API_BASE` in the same order as other OpenAI calls
- **Validation before any upstream call**: a missing `state`, a malformed question, more than 128 questions, a single-option question or a body that is not JSON returns 400. Image input sent to a provider that only accepts text returns a clear 400
- **Spend tracking**: per-token pricing from the cost map, with cached and cache-write input tokens billed at their own rates
- **Health checks**: `/health` probes `mode: evaluation` deployments with a one-question Decisions call, and `health_check_params` can override its `state` and `questions`
- **Guardrails**: OpenAI-format responses include `guardrail_information` when the caller sets `include_guardrail_response`
- **Pass-through**: an existing `pass_through_endpoints` entry at `/v1/decisions` keeps answering that path, while `/decisions` serves the native route

#### New Model Support (6 new models)

| Provider | Model | Context Window | Input ($/1M tokens) | Output ($/1M tokens) |
| --- | --- | --- | --- | --- |
| Perplexity | `perplexity/pplx-decider-v1-27b` | 262K | $0.04 | $0.00 |
| Cloudflare | `cloudflare/clef` | 65K | $0.24 | $0.00 |
| Cloudflare | `cloudflare/clef-flash` | 65K | $0.09 | $0.00 |
| Cloudflare | `cloudflare/@cf/cloudflare/clef` | 65K | $0.24 | $0.00 |
| Cloudflare | `cloudflare/@cf/cloudflare/clef-flash` | 65K | $0.09 | $0.00 |
| Strands Decider | `strands_decider/strands-decider-2B-hobson-v19` | - | $0.00 (self-hosted) | $0.00 |

The existing `gpt-6-luna` entry now lists `/v1/decisions` among its supported endpoints

### What's Changed

- Add a unified `/v1/decisions` endpoint for System One-compatible providers - [PR #44236](https://github.com/BerriAI/litellm/pull/44236)
- Serve the System One format at `/v1/systemone` and OpenAI's format at `/v1/decisions` - [PR #45184](https://github.com/BerriAI/litellm/pull/45184)
- Add OpenAI as a Decisions provider behind a shared decisions format - [PR #45214](https://github.com/BerriAI/litellm/pull/45214)

All three are backported in [PR #45189](https://github.com/BerriAI/litellm/pull/45189)

## Full Changelog

https://github.com/BerriAI/litellm/compare/v1.105.0-rc.2...v1.105.0-rc.3

## Related pages

- [v1.105.0rc1](https://docs.litellm.ai/release_notes/v1.105.0rc1/v1-105-0-rc-1.md)
- [v1.104.2](https://docs.litellm.ai/release_notes/v1.104.2/v1-104-2.md)