PiAPI API pricing: from $0.0051-$0.161/success

Set up

Set up PiAPI in Claude Code, Codex or any agent

1Give this to your agent

in your agent's chat

set up treg — https://treg.to/llms.txt

2Or add the MCP server yourself

claude code

claude mcp add --transport http treg https://treg.to/mcp

ChatGPT

Settings → Connectors → add https://treg.to/mcp as a custom connector.

Grok bot

Add https://treg.to/mcp as a remote MCP connector in its tool settings.

Claude Desktop

Settings → Connectors → Add custom connector → https://treg.to/mcp.

Cursor, Codex, any MCP client

Point it at https://treg.to/mcp (HTTP transport).

CLI

curl -fsSL https://treg.to/install.sh | sh

Plain HTTP

LangChain, CrewAI or any code: /call/<tool-id> with a Bearer token. No SDK.

Why treg.to

Why call PiAPI through treg.to

No PiAPI signup

Eligible tools run on treg.to's key, metered per call from a prepaid balance.

Price before the call

The provider's own rate, $0.000 markup. How billing works.

No subscription, no seats

Platform calls are charged per call. $1.00 free per new team, no card to start.

Your own PiAPI key is free

Switch by changing a word

Another provider is a different word in the prompt, not a new integration.

One key, the whole catalog

The same token calls OpenRouter, reAPI, Replicate and 2,907 other tools.

The shelf

All 6 PiAPI tools

Image generation · 3 tools

Google: Gemini 3 Pro Image (Nano Banana Pro) · Platform + BYOK · live-verified · $0.105-$0.181/success
Generate or edit an image with Gemini 3 Pro Image at 1K, 2K or 4K from a prompt and up to 14 reference images. piapi.image-gen.gemini-3-pro-image
One image per task. Live 2026-09-14: a 1K image consumed 1,050,000 points ($0.105); 2K and 4K are documented only.
OpenAI: GPT Image 2 · Platform + BYOK · live-verified · $0.0064-$0.169/success
Generate images with GPT Image 2 (medium quality only) through PiAPI's OpenAI-shaped async route. piapi.image-gen.gpt-image-2
Token-metered upstream ($30 per 1M output tokens, passed through at cost); medium is the only quality. Cells are the points PiAPI consumed live on 2026-09-14 (63,450 for 1024x1024; 420,350 for either non-square size, at 10,000,000 points per dollar), rounded up. The docs' ~$0.032 figure for a square image was not what the meter charged.
OpenAI: GPT Image 2.5 (Flare / Sunburst) · Platform + BYOK · live-verified · $0.0051-$0.161/success
Generate images with GPT Image 2.5 through PiAPI's OpenAI-shaped async route, metered by quality and size. piapi.image-gen.gpt-image-2-5
Token-metered upstream (OpenAI's $3.75 / $22.50 per 1M input / output tokens, passed through at cost). Cells are the points PiAPI consumed live on 2026-09-14 (50,663 / 105,338 / 401,663 points at 10,000,000 points per dollar for low / medium / high), rounded up to the next hundredth of a cent; Flare and Sunburst meter identically. A very long prompt adds a fraction of a cent the table does not carry.

Video generation · 3 tools

Check a PiAPI task · Platform + BYOK · helper · live-verified · free
Read a task's status (Pending, Processing, Completed, Failed, Staged) and its output URLs. piapi.task.get
Task status checks are free.
ByteDance: Seedance 2.5 · Platform + BYOK · live-verified · $0.75-$24/success
Create a Seedance 2.5 video task from a prompt with optional image, video and audio references (default content filter). piapi.video-gen.seedance-2-5
Billed per second of output at the resolution's rate. Live 2026-09-14: a 5 s 480p clip consumed 7,500,000 points ($0.75, at 10,000,000 points per dollar); 720p and 1080p are documented only.
ByteDance: Seedance 2.5 (Less Restriction) · Platform + BYOK · live-verified · $0.825-$26.4/success
Less Restriction: the same Seedance 2.5 model on PiAPI's relaxed filter, where a real person's face may drive the video; the default row refuses real faces. Face images must first pass the connected account's Private Asset Library review; other references are used as-is. 10% dearer than the default row. piapi.video-gen.seedance-2-5.less-restriction
10% over the default-filter row at every resolution. Live 2026-09-14: a 5 s 480p clip from one reference image consumed 8,250,000 points ($0.825); 720p and 1080p are documented only.

Reliability badges come from live traffic through treg.to. Live-verified means a successful provider check; it does not verify every request variant.

Questions

Before you start

Do I need a PiAPI account?

No. Eligible tools run on treg.to's key and the call is metered from your team's prepaid balance, priced up front. If your team registers its own PiAPI key, that key always wins and those calls are never metered.

What does a call cost?

Each tool on this page shows its rate; the cheapest is $0.0051-$0.161/success. The rate is the provider's own and treg.to adds no markup; it is billed per call from a prepaid balance. It is not PiAPI's subscription pricing, which is on their own site. New teams start with $1.00 of free credit.

How do I add PiAPI to Claude Code?

Run: claude mcp add --transport http treg https://treg.to/mcp. One MCP server carries PiAPI and the rest of the catalog.

Which frameworks does it work with?

Anything that speaks MCP (Claude Code, Claude Desktop, ChatGPT, Codex, Cursor, Grok) and anything that can make an HTTP request (LangChain, CrewAI, LlamaIndex, plain code).

Is this the official PiAPI MCP server?

No. treg.to serves PiAPI's real API through its own metered proxy: the request is the provider's own, the credential is injected server-side, and the answer is relayed verbatim. The official PiAPI channels are linked above.

PiAPI: 6 tools from $0.0051-$0.161/success

PiAPI, one prompt away