Inference Logoinference shell

BAGEL Image Understanding & Generation

A tool that combines understanding and creation capabilities, notable for its high-quality image generation and advanced image editing using simple language.

reasoningtext to imageimage to imageimage editing
run in browser →run via API
run with your agent
# install belt
$curl -fsSL https://cli.inference.sh | sh
# sign in (or set INFERENCE_API_KEY on a headless machine)
$belt login
# view schema & details
$belt app get infsh/bagel
# run
$belt app run infsh/bagel --input '{"system_prompt":"you are a helpful assistant that can answer questions and help with tasks.","context":[{"content":[{"text":"What is the capital of France?","type":"text"}],"role":"user"},{"content":[{"text":"The capital of France is Paris.","type":"text"}],"role":"assistant"}],"text":"write a haiku about artificial general intelligence"}'
maker
infsh
category
image
input → output
text → image, image → image
price
GPU time, billed per millisecond
runs via
api, javascript and python sdk, belt cli, mcp
updated
2026-07-28

api reference

about

a tool that combines understanding and creation capabilities, notable for its high-quality image generation and advanced image editing using simple language.

1. calling the api

install the client

the client provides a convenient way to interact with the api.

bash
1pip install inferencesh

setup your api key

set INFERENCE_API_KEY as an environment variable. get your key from settings → api keys.

bash
1export INFERENCE_API_KEY="inf_your_key"

run and get result

submit a request and wait for the final result. best for batch processing or when you don't need progress updates.

python
1import os2from inferencesh import inference34client = inference(api_key=os.environ["INFERENCE_API_KEY"])567result = client.run({8        "app": "infsh/bagel",9        "input": {10            "system_prompt": "you are a helpful assistant that can answer questions and help with tasks.",11            "context": [12                {13                    "content": [14                        {15                            "text": "What is the capital of France?",16                            "type": "text"17                        }18                    ],19                    "role": "user"20                },21                {22                    "content": [23                        {24                            "text": "The capital of France is Paris.",25                            "type": "text"26                        }27                    ],28                    "role": "assistant"29                }30            ],31            "text": "write a haiku about artificial general intelligence"32        }33    })3435print(result["output"])

stream live updates

get real-time progress updates as the task runs. ideal for showing progress bars, partial results, or long-running tasks.

python
1import os2from inferencesh import inference34client = inference(api_key=os.environ["INFERENCE_API_KEY"])567# stream=True yields updates as they arrive8for update in client.run({9        "app": "infsh/bagel",10        "input": {11            "system_prompt": "you are a helpful assistant that can answer questions and help with tasks.",12            "context": [13                {14                    "content": [15                        {16                            "text": "What is the capital of France?",17                            "type": "text"18                        }19                    ],20                    "role": "user"21                },22                {23                    "content": [24                        {25                            "text": "The capital of France is Paris.",26                            "type": "text"27                        }28                    ],29                    "role": "assistant"30                }31            ],32            "text": "write a haiku about artificial general intelligence"33        }34    }, stream=True):35    if update.get("progress"):36        print(f"progress: {update['progress']}%")37    if update.get("output"):38        print(f"output: {update['output']}")

2. authentication

the api uses api keys for authentication. see the authentication docs for detailed setup instructions.

3. files

file inputs are automatically handled by the sdk. you can pass local paths, urls, or base64 data.

automatic upload

the python sdk automatically detects local file paths and uploads them. urls are passed through as-is.

python
1# local file paths are automatically uploaded2result = client.run({3    "app": "infsh/bagel",4    "input": {5        "image": "/path/to/local/image.png",  # detected & uploaded6        "audio": "https://example.com/audio.mp3",  # url passed through7    }8})

manual upload

you can also upload files manually and use the returned url.

python
1# upload and get a hosted URL2file = client.files.upload("/path/to/file.png")3print(file.uri)  # https://cloud.inference.sh/...

4. webhooks

get notified when a task completes by providing a webhook url. when the task reaches a terminal state (completed, failed, or cancelled), a POST request is sent to your url with the task result.

python
1result = client.run({2    "app": "infsh/bagel",3    "input": {4        "system_prompt": "you are a helpful assistant that can answer questions and help with tasks.",5        "context": [6            {7                "content": [8                    {9                        "text": "What is the capital of France?",10                        "type": "text"11                    }12                ],13                "role": "user"14            },15            {16                "content": [17                    {18                        "text": "The capital of France is Paris.",19                        "type": "text"20                    }21                ],22                "role": "assistant"23            }24        ],25        "text": "write a haiku about artificial general intelligence"26    },27    "webhook": "https://your-server.com/webhook"28}, wait=False)

webhook payload

your endpoint receives a JSON POST with the task result:

json
1{2  "event": "task.completed",3  "timestamp": "2024-01-15T10:30:05Z",4  "data": {5    "id": "task_abc123",6    "short_id": "abc123",7    "status": 10,8    "status_text": "completed",9    "output": { ... },10    "created_at": "2024-01-15T10:30:00Z",11    "updated_at": "2024-01-15T10:30:05Z"12  }13}
eventstringtask.completed, task.failed or task.cancelled
timestampstringiso timestamp of the delivery
data.idstringtask id
data.short_idstringshort task id
data.statusnumberterminal status (10=completed, 11=failed, 12=cancelled)
data.status_textstringthe status as a word: completed, failed or cancelled
data.outputobjecttask output (when completed)
data.errorstringerror message (when failed, omitted otherwise)
data.session_idstringsession id (when the task ran in a session, omitted otherwise)
data.created_atstringiso timestamp
data.updated_atstringiso timestamp

5. schema

input

thinking_enabledboolean

enable thinking mode for enhanced reasoning

default: false
analysis_modeboolean

enable understanding mode for image analysis only

default: false
imagesarray

the images to use for the model

system_promptstring

the system prompt to use for the model

default: "you are a helpful assistant that can answer questions and help with tasks."example: "you are a helpful assistant that can answer questions and help with tasks."
contextarray

the context to use for the model

default: []example: [{"content":[{"text":"What is the capital of France?","type":"text"}],"role":"user"},{"content":[{"text":"The capital of France is Paris.","type":"text"}],"role":"assistant"}]
rolestring

the role of the input text

default: "user"
options:"user""assistant""system""tool"
textstring*

the input text to use for the model

example: "write a haiku about artificial general intelligence"
temperaturenumber

temperature

default: 0.7min:0max:1
top_pnumber

top p

default: 0.95min:0max:1
context_sizeinteger

context size

default: 4096
max_tokensinteger

max tokens

default: 64000

output

imagestring(file)

generated or processed image

reasoningstring

thinking content

responsestring

generated text response (thinking or understanding)

other image apps

Gemini Nano Banana 2.1

google/gemini-nano-banana-2-1

$1.50/M input, $0.0336–$0.1134/image (1K–4K), +$0.014/query Google Search

grok-imagine-image-2-0

xai/grok-imagine-image-2-0

$0.04/img low 1k, $0.06/img low 2k, $0.06/img med 1k, $0.08/img med 2k; $0.01/source img (edits)

GPT Image 2.5 Sunburst

openai/gpt-image-2-5-sunburst

$0.006-$0.21 per 1024x1024 image: low $0.006, medium $0.013, high $0.053, xhigh $0.094, max $0.21. Larger sizes cost more; edits add ~$0.008 per reference image.

GPT Image 2.5 Flare

openai/gpt-image-2-5-flare

$0.006-$0.21/image at 1024x1024: low $0.0059, medium $0.0132, high $0.0527, xhigh $0.0937, max $0.2107. Larger sizes cost more; edits add ~$0.0083 per reference image.

Image Resize & Crop

eval/resize

from $0.00000277/sec

more from infsh

HTTP Request

infsh/http

from $0.00000277/sec

DramaBox Text-to-Speech

infsh/dramabox

GPU time, billed per millisecond

Markdown to Word Converter

infsh/markdown-to-word

from $0.00000277/sec

Harrier OSS v1 Text Embeddings

infsh/harrier-oss-v1

GPU time, billed per millisecond

ready to run BAGEL Image Understanding & Generation?

try in browserbrowse all apps

we use cookies

we use cookies to ensure you get the best experience on our website. for more information on how we use cookies, please see our cookie policy.

by clicking "accept", you agree to our use of cookies.
learn more.