# Lock-in agent planner models

The Lock-in agent plans your work with a planner model you can switch: Auto plus 30 others, priced per message from 4 to 369 credits. Auto is free for 50 messages per UTC day.

Planner models are the chat models behind the Lock-in agent. They read your brief, ask questions and propose image, video and audio generations. The image, video and audio models make the media, and each proposal shows its own price. There are 31 choices: Auto plus 30 models.

## Auto

- Auto is the default. It picks a model for each message, so you do not have to.
- Auto is free for the first 50 messages per user per UTC day. After that each message costs 5 credits.
- A trivial turn, such as a plain acknowledgement, goes to GPT-6 Luna. Most messages go to Gemini 3.8 Flash. A message with an image you attached goes to Claude Opus 5.5.

## How a message is priced

- Each planner model has a price per message, from 4 to 369 credits. The tables below list every one.
- If a model fails before it starts to answer, the agent tries a fallback model. A fallback is charged at the cheaper of the two prices.
- The images, videos and audio the agent proposes are priced separately, at the image, video and audio model prices. In Ask mode the agent shows each proposal with its price and waits for your approval. In Auto mode it runs without approval. This mode is separate from the Auto model.

## Featured models

The model menu shows a Featured list first. It holds Auto and 12 models.

| Model | Maker | Reads images | Credits per message |
| --- | --- | --- | --- |
| Auto | UOVA | Yes | Free for 50 messages per user per UTC day, then 5 |
| GPT-6 Astra | OpenAI | Yes | 369 |
| Claude Opus 5.5 | Anthropic | Yes | 146 |
| GPT-6 Sol | OpenAI | Yes | 74 |
| Claude Fable 5.1 | Anthropic | Yes | 362 |
| Grok 4.7 | xAI | Yes | 40 |
| GLM-5.3 Flash | Z.ai | Yes | 5 |
| Claude Fable 5 | Anthropic | Yes | 369 |
| GPT-6 Luna | OpenAI | Yes | 4 |
| GLM-5.3 | Z.ai | No, text only | 45 |
| Kimi K3 | Moonshot | Yes | 105 |
| DeepSeek V4.1 Flash | DeepSeek | Yes | 10 |
| Gemini 3.8 Flash | Google | Yes | 27 |

## All other models

The All models list holds 18 more.

| Model | Maker | Reads images | Credits per message |
| --- | --- | --- | --- |
| GPT-5.6 Sol | OpenAI | Yes | 148 |
| Claude Opus 5 | Anthropic | Yes | 185 |
| Grok 4.6 | xAI | Yes | 66 |
| DeepSeek V4 Flash | DeepSeek | No, text only | 4 |
| Claude Opus 4.8 | Anthropic | Yes | 185 |
| Muse Spark 1.3 | Meta | Yes | 41 |
| Seed 2.1 Turbo | ByteDance | Yes | 18 |
| Claude Sonnet 5 | Anthropic | Yes | 74 |
| Grok 4.5 | xAI | Yes | 64 |
| GPT-5.6 Luna | OpenAI | Yes | 8 |
| GLM-5.2 | Z.ai | No, text only | 27 |
| Claude Opus 4.6 | Anthropic | Yes | 185 |
| Gemini 3.1 Pro | Google | Yes | 73 |
| GPT-5.5 | OpenAI | Yes | 182 |
| Claude Sonnet 4.6 | Anthropic | Yes | 111 |
| DeepSeek V4 Pro | DeepSeek | No, text only | 41 |
| Gemini 3 Flash | Google | Yes | 19 |
| GPT-5.2 | OpenAI | Yes | 70 |

## Images and text-only models

Four models read text only: GLM-5.3, DeepSeek V4 Flash, GLM-5.2 and DeepSeek V4 Pro. If you send one of them a message that carries an image, the agent answers with Auto's image model, Claude Opus 5.5, and tells you in one line.

## Limits per message

- Up to 6 images attached.
- Up to 8,000 characters of text.
- Up to 4 proposals, each making up to 4 outputs.

## Common questions

**How many Auto messages are free each day?**

50 per user per UTC day. After the 50th, each Auto message costs 5 credits until the next UTC day.

**Which planner models cost the least?**

GPT-6 Luna and DeepSeek V4 Flash cost 4 credits per message. DeepSeek V4 Flash reads text only. GLM-5.3 Flash costs 5 credits and reads images.

**Does the planner model make my images?**

No. The planner model plans and proposes. The image, video and audio models listed on [Image models](https://www.uova.com/help/models/image), [Video models](https://www.uova.com/help/models/video) and [Audio models](https://www.uova.com/help/models/audio) make the media.

Source: https://www.uova.com/help/models/chat