Upscale diffusion
NativeImageupscale.diffusionNobody serves this model right now. The market reopens when a seller comes online.
Performance
Measured from real, settled requests over the last 30 days. No synthetic benchmarks.
Availability
Provider availability polled every minute by the platform itself, replayed hour by hour.
0% online over the last 3 days
API
The API is OpenAI-compatible and uses the same price and sellers as shown here.
curl https://api.openmayhem.ai/v1/workflows \
-H "Authorization: Bearer $OPENMAYHEM_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "upscale.diffusion",
"workflow": {"<node-id>": {"class_type": "<allowed-node>", "inputs": {}}}
}'MCP
OpenMayhem runs its own MCP server, so Claude, Codex, Cursor and other AI apps can call this model and pay from your credit.
claude mcp add --transport http openmayhem \ https://mcp.openmayhem.ai/mcp \ --header "X-OpenMayhem-API-Key: $OPENMAYHEM_API_KEY"
Then ask your app, for example: "Use OpenMayhem Upscale diffusion to answer this." It calls generate_image with upscale.diffusion. Apps that support sign-in connect without a key, and you approve spending limits once.
Full request reference
POST /v1/workflowsWORKFLOWS · mayhem_comfy_workflows +
Working request
curl https://api.openmayhem.ai/v1/workflows \
-H "Authorization: Bearer $OPENMAYHEM_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $(uuidgen)" \
-d '{
"model": "upscale.diffusion",
"workflow": {},
"input_files": [
{
"filename": "refs/input.png",
"kind": "image",
"content_type": "image/png",
"encoding": "base64",
"data": "<BASE64_IMAGE>"
}
],
"response_format": "artifact",
"timeout_ms": 300000
}'Parameters
| Parameter | Required | Type | Default | Allowed values | Limits and shape |
|---|---|---|---|---|---|
input_files | Required | array | — | — | min items 1; max items 4 [{"filename":"refs/input.png","kind":"image","content_type":"image/png","encoding":"base64","data":"<BASE64_IMAGE>"}] |
model | Required | string | — | — | — |
workflow | Required | object | — | — | {"node-id":{"class_type":"AllowedNode","inputs":{}}} |
outcome_class | Optional | string | — | — | min length 1; max length 128 |
response_format | Optional | string | artifact | artifact, json | {"type":"json_object"} |
runtime_id | Optional | string | — | — | min length 1; max length 128 |
timeout_ms | Optional | integer | 300000 | — | min 1000; max 900000 |
user | Optional | string | — | — | min length 1; max length 512 |
Input shapes
{
"model": "upscale.diffusion",
"input_files": [
{
"filename": "refs/input.png",
"kind": "image",
"content_type": "image/png",
"encoding": "base64",
"data": "<BASE64_IMAGE>"
}
]
}Workflow surface
Use the complete starter. Node roles, classes, inputs, presets and add-ons below are the values admitted for this workflow.
Use this model through MCP
These calls use this model’s exact live request contract. First estimate the unchanged request, then pass the returned estimate ID and maximum cost to the paid tool. Read the complete MCP guide.
run_workflowWORKFLOWS +
Send this object to estimate_request. Review the result, then send the next object to the named paid tool without changing request.
{
"endpoint": "WORKFLOWS",
"model": "upscale.diffusion",
"request": {
"workflow": {},
"input_files": [
{
"filename": "refs/input.png",
"kind": "image",
"content_type": "image/png",
"encoding": "base64",
"data": "<BASE64_IMAGE>"
}
],
"response_format": "artifact",
"timeout_ms": 300000
}
}{
"model": "upscale.diffusion",
"request": {
"workflow": {},
"input_files": [
{
"filename": "refs/input.png",
"kind": "image",
"content_type": "image/png",
"encoding": "base64",
"data": "<BASE64_IMAGE>"
}
],
"response_format": "artifact",
"timeout_ms": 300000
},
"estimate_id": "<estimate_id from estimate_request>",
"max_cost_usd": "<maximum_cost_usd from estimate_request>"
}This model returns a job. Poll it until completion, then create a short-lived artifact link.
{
"job_id": "<job_id from the paid tool>"
}{
"artifact_id": "<artifact_id from get_job>",
"download": false
}Agent drop-ins
This is a media endpoint, so it cannot replace the text model that drives a coding agent. Call the request above from the agent’s HTTP or tool layer.
About this model
An image upscaling workflow; convolutional variants are fast, the diffusion variant re-synthesises detail.
- Maker
- Network workflow
- Family
- Upscale diffusion
How it is served
- Registered tiers
- T4
- Live tiers
- Not advertised
- Quantizations
- unknown
- Live routes
- 0
- Live providers
- 0
- Prompt processing
- Provider-visible
- Context capacity
- 1
Routing providers
Accepted request parameters
From the network's signed model contract; anything else is rejected before a provider runs.
input_filesmodeloutcome_classresponse_formatruntime_idtimeout_msuserworkflowPrice versions
| Version | From | Until |
|---|---|---|
| v6 | 2026-09-22 17:49 UTC | current |
| v5 | 2026-09-16 22:12 UTC | 2026-09-22 17:49 UTC |
| v4 | 2026-09-14 06:13 UTC | 2026-09-16 22:12 UTC |
| v3 | 2026-09-14 01:36 UTC | 2026-09-14 06:13 UTC |
| v2 | 2026-09-02 17:01 UTC | 2026-09-14 01:36 UTC |
| v1 | 2026-08-30 23:46 UTC | 2026-09-02 17:01 UTC |
Questions and answers
What does Upscale diffusion cost on OpenMayhem?
Upscale diffusion currently costs $0.045 per 1000 megapixel step. There is no subscription: you buy credit and every request settles against a signed receipt for exactly what it used. Prices are versioned, and the full price history is published on this page.
Is Upscale diffusion available right now?
No providers are online for Upscale diffusion at this moment. Availability is polled every minute, and requests are rejected upfront rather than queued indefinitely.
How do I use Upscale diffusion?
Send an OpenAI-compatible request to https://api.openmayhem.ai/v1/workflows with "model": "upscale.diffusion" and your API key. Existing OpenAI SDKs work by changing the base URL. Create an account to get a key and starting credit.
Are my prompts stored?
OpenMayhem does not persist prompt or completion text. Requests are proxied to the serving provider and only billing metadata is retained: usage counts, the price applied, and the provider's signed receipt. Generated media files are stored temporarily so you can download them, then expire. The provider that serves a request necessarily processes its content.