# CherryText 1.0 Flash

> Fast, low-cost language model with a 1M-token context window, tool calls, and reasoning.

> Sakuranbo AI is in preview.

- Model ID: `sakuranbo/cherrytext-1.0-flash`
- Vendor: Sakuranbo AI
- Modality: Text
- Tasks: Chat
- Upstream status: GA (upstream)
- Last updated: 2026-10-29
- HTML page: https://sakuranbo.ai/models/sakuranbo/cherrytext-1.0-flash/
- llms.txt: https://sakuranbo.ai/models/sakuranbo/cherrytext-1.0-flash/llms.txt

## Description

CherryText 1.0 Flash is a fast, low-cost language model for chat, coding, and agent workloads, with a 1M-token context window and up to 131,072 output tokens. Call it through the OpenAI-compatible Chat Completions endpoint with streaming, tool calls, and JSON output. The model reasons before it answers and returns the reasoning in the reasoning_content field.

## Capabilities

- Input: Text
- Output: Text
- Context length: 1,000,000 tokens
- Max output: 131,072 tokens
- Streaming: Yes
- Asynchronous job: No

## Parameters

| Name | Type | Required | Default | Constraints | Description |
| --- | --- | --- | --- | --- | --- |
| `messages` | object[] | Yes |  |  | Conversation messages in the OpenAI Chat Completions format. |
| `max_tokens` | integer | No | `4096` | 1 to 131,072 | Maximum number of tokens to generate, including the tokens the model spends on reasoning. A value that is too small can end the response before the answer is written. The balance hold is based on this value. |
| `temperature` | number | No | `1` | 0 to 2 | Sampling temperature. |
| `stream` | boolean | No | `false` |  | Stream tokens as server-sent events. |

## Pricing

Prices are in USD and include our margin. Failed, cancelled, expired, and policy-rejected runs are not billed. For chat, once the model has started generating output (including reasoning output), the output it generates is billed, even if the stream breaks or you close the connection.

| Dimension | Unit | Condition | Price |
| --- | --- | --- | --- |
| Input tokens | per 1M tokens | Default | $0.4125 |
| Cached input tokens | per 1M tokens | Default | $0.0104 |
| Output tokens | per 1M tokens | Default | $1.65 |

### Price examples

Examples use the listed prices. The server calculates the final charge from actual usage.

- Default run: $0.00676 (empty input, up to 4,096 output tokens)
- 1,000 input and 1,000 output tokens: $0.002063
- 10,000 input and 1,000 output tokens: $0.005775
- 100,000 input and 4,096 output tokens: $0.048009

## API

### Estimate the price

POST /v1/estimate takes a JSON body ({ model, input }) and returns the price for that input before you run it. This endpoint does not need an API key.

```bash
curl -X POST "https://api.sakuranbo.ai/v1/estimate" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "sakuranbo/cherrytext-1.0-flash",
  "input": {
    "messages": [
      {
        "role": "system",
        "content": "You are a concise assistant."
      },
      {
        "role": "user",
        "content": "Summarize the benefits of prepaid API credits in two lines."
      }
    ],
    "max_tokens": 2048
  }
}'
```

### Generate

Send generation requests to /v1/chat/completions. The request below uses an example input from the catalog for this model.

```bash
curl -X POST "https://api.sakuranbo.ai/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $SAKURANBO_API_KEY" \
  -d '{
  "model": "sakuranbo/cherrytext-1.0-flash",
  "messages": [
    {
      "role": "system",
      "content": "You are a concise assistant."
    },
    {
      "role": "user",
      "content": "Summarize the benefits of prepaid API credits in two lines."
    }
  ],
  "max_tokens": 2048
}'
```

[Read the Chat guide](https://sakuranbo.ai/docs/chat/)

## Data policy

- Outputs are returned in the response and are not stored.
- Prompts are not stored. Only request metadata is kept.
- Your request is forwarded to an upstream provider that runs the model.

## Related models

- [Cherry 1.0 Pro](https://sakuranbo.ai/models/sakuranbo/cherry-1.0-pro.md): Text- and image-to-video generation from 2 to 30 seconds at 1080p, 2K, or 4K with audio.
- [CherryImage 1.0 Pro](https://sakuranbo.ai/models/sakuranbo/cherryimage-1.0-pro.md): High-fidelity text-to-image and reference-based image editing up to 2K resolution.

Source: https://sakuranbo.ai/models/sakuranbo/cherrytext-1.0-flash/