v1.0 · public beta · 12 models · 1 endpoint

Design generation as an API. And an MCP.

Twelve current models, one endpoint, one brand kit. Drop on-brand design generation into Claude, Cursor, your own agent, or your app. Every aspect ratio. Every output format. One monthly plan covers every model — no per-model billing math, no surprise invoices when Veo gets called instead of Gemini.

Get an API key Read the MCP docs

Start free to explore · API + MCP + video included on Pro ($59/mo)

Compatible with

CLClaude Desktop CUCursor CNCline n8n8n ZPZapier MKMake

JavaScript cURL Python MCP

import { Ohoney } from "@ohoney/sdk";

const client = new Ohoney({ apiKey: process.env.OHONEY_API_KEY });

const { designs } = await client.generate({
  niche: "cozy autumn florals",
  brandKit: "kit_4xQ2",
  format: "1:1",
  count: 12,
});

for (const d of designs) console.log(d.id, d.url);

→ 12 on-brand designs · transparent PNG + print-ready upscale included · counts as 12 against your monthly quota

Why Ohoney for builders

The design layer your raw-model API doesn't have.

Brand-kit aware by default

Pass a brand kit ID, get output that respects your colors, fonts, voice, and product context. Raw model APIs don't know what your brand is — Ohoney does, because that's the layer.

∞

One endpoint, every model

Image, video, vector, sharp text, background removal, upscaling — every leading model behind a single POST /v1/generate. We route by task. New models added the moment they launch.

⌖

Multi-format in one call

Ask for 1:1, 9:16, 16:9, 4:5 in one request — get every aspect ratio back. No re-rolls. No per-format pricing. The variant matrix is the default.

⟁

Predictable monthly billing

Monthly plan with a design quota. We absorb the per-model upstream cost — Veo costs 30× what Gemini costs and you never see it. Pro ($59/mo) includes 700 designs, video generation, API, MCP, and bulk batches. Top-up packs handle overflow.

Native MCP

Drop Ohoney into any agent.

First-class Model Context Protocol server. Claude Desktop, Cursor, Cline, custom agents — wire Ohoney in once and your agent can generate on-brand designs as a tool call. Same brand kits, same models, same pricing.

claude_desktop_config.json cursor.json manual

{
  "mcpServers": {
    "ohoney": {
      "command": "npx",
      "args": ["-y", "@ohoney/mcp-server"],
      "env": {
        "OHONEY_API_KEY": "ok_live_..."
      }
    }
  }
}

→ Adds ohoney.generate, ohoney.recolor, ohoney.upscale, ohoney.video as agent tools

Tools exposed to agents

ohoney.generateCreate on-brand designs from a niche keyword + brand kit
ohoney.recolorFan one design across every color variant your buyer might pick
ohoney.upscalePrint-ready upscale, alpha-preserving, lossless
ohoney.background_removeClean transparent cutouts for POD platforms
ohoney.videoVeo-quality video from a design or product prompt
ohoney.list_brand_kitsEnumerate brand kits the API key has access to

Built for

The people building with AI under the hood.

⌥

Agent + MCP builders

Give Claude, Cursor, or your custom agent the ability to generate on-brand visuals as a tool call. Brand-kit-aware. No per-model billing the agent has to reason about. Output sized for the channel.

▤

SaaS founders building design features

Skip building model-routing, brand-application, and output-format logic yourself. Hit one endpoint, get on-brand designs back. Your users see the output; you see the cost line.

⊞

Workflow + automation engineers

Wire Ohoney into n8n, Make, Zapier, or your own job runner. Batch-generate a season's worth of POD designs overnight. Webhook fires on completion. Cost ceiling per run.

⌗

AI app shops + platform teams

White-label the entire design layer for your own platform. SSO, per-tenant brand kits, audit log, dedicated CSM. SOC 2 in progress.

Every model · Always current

Twelve current models. Every leading lab. One endpoint.

Image, video, vector, sharp text, background removal, upscaling. We route to the right model per task and absorb the upstream pricing. New models added the moment they launch.

Gemini 2.5 Flash Imageimage · standard

Recraft v3image · vector

Imagen 4image · photo

Ideogram v3image · typography

GPT Image 1image · OpenAI

Runway Gen-3 Turbovideo · standard

Seedance 2.0video · mid

Veo 3video · premium

Pollinations Fluximage · slow

WAN 2.5video · slow

BiRefNetbg-remove

Real-ESRGANupscale

Pricing

One monthly plan. Every model in scope.

Designs count against your monthly quota. The upstream cost (Gemini vs Veo vs Recraft) is our problem — your invoice stays the same shape. API and MCP unlock on Pro. Top-up packs cover spikes.

Free $0 25 designs / mo · explore the product · designs train the public library · no API

Creator $19 / mo 200 designs / mo · commercial license · private vault · marketplace manifests · no API

Pro · API + MCP $59 / mo 700 designs / mo · video generation (Veo, Runway) · API + MCP access · priority queue · bulk up to 100 · top-up packs

Full pricing comparison → Team plans →

Start building with Ohoney.

One endpoint, twelve models, native MCP, brand-kit-aware output. Start free, upgrade to Pro when you need the API.

Get an API key Read MCP docs