Every model,one API key.
genmux API is one image and chat models API: one key, one balance, every model.
Pay per image or per token, in credits. OpenAI-compatible chat, async image jobs with webhooks, failed jobs refunded.
Per image or per 1M tokens, at the Scale pack rate. Failed image jobs are refunded; chat is billed on actual usage.
Popular and best-value models
The models people start with, plus the ones with the biggest verified saving against their maker's list price.
Savings compare our price at the Scale pack rate with the maker's published list price on a 75 % input / 25 % output mix; list prices checked September 6, 2026.
Works with your coding agent
One key, one OpenAI-compatible endpoint with tool calling; paste a verified config and pick a model with tools: true.
OpenCode · Aider · Cline · Roo Code · Continue
Tool calling for coding agents
POST /v1/chat/completions with OpenAI function calling — tools, tool_choice, streamed tool_calls — so OpenCode, Aider, Cline and Continue run on DeepSeek, Qwen3 and Doubao unchanged.
Async image jobs with webhooks
Submit a generation, get a job id back instantly, and get notified the moment it finishes — no polling required.
Exactly-once billing, one balance
Images are charged when the job succeeds and refunded on failure; chat is settled on the tokens actually used. One credit balance for both.
Keys, limits, allowlists
Scope API keys per project with request-per-second limits and CIDR allowlists you control from the dashboard.
Two requests.
One image — or one reply.
Submit a job, poll for the result. Every response below is real — copy the commands with your key and they run as-is.
One request, streamed back token by token. The last chunk carries the token usage and the cost in credits — the same shape the OpenAI SDK already understands.
Full API reference →01 · Submit the jobt + 0 ms
curl -s -X POST https://api.genmux.tech/v1/generations \ -H "Authorization: Bearer $API_KEY" \ -H "Idempotency-Key: $(uuidgen)" \ -H "Content-Type: application/json" \ -d '{"model":"nanobanana","prompt":"a red fox in snow","resolution":"1K","ratio":"1:1"}'02 · Accepted — you hold the job idt + 41 ms
202 Accepted{ "id": "0d8b6a3e-2f1c-4a5b-9c1d-7e2f3a4b5c6d", "model": "nanobanana", "status": "queued", "cost": 2 }03 · Poll, or wait for the webhookt + 1 s
curl -s https://api.genmux.tech/v1/generations/0d8b6a3e-2f1c-4a5b-9c1d-7e2f3a4b5c6d \ -H "Authorization: Bearer $API_KEY"04 · Done — the image is at the URLt + 52 s
200 OK{ "id": "0d8b6a3e-2f1c-4a5b-9c1d-7e2f3a4b5c6d", "model": "nanobanana", "status": "succeeded", "cost": 2, "images": [{ "url": "https://…/0.png?X-Amz-Expires=900", "width": 1024, "height": 1024 }]}
Send a message
curl -N -X POST https://api.genmux.tech/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"qwen3.8-max","stream":true,
"messages":[{"role":"user","content":"Explain idempotency keys in two sentences."}]}'The stream
data: {"id":"7c1e…","object":"chat.completion.chunk","model":"qwen3.8-max","choices":[{"index":0,"delta":{"role":"assistant","content":"An"},"finish_reason":null}]}data: {"id":"7c1e…","choices":[{"index":0,"delta":{"content":" idempotency key"},"finish_reason":null}]}data: {"id":"7c1e…","choices":[{"index":0,"delta":{"content":" lets a client retry"},"finish_reason":null}]}data: {"id":"7c1e…","choices":[{"index":0,"delta":{"content":" a request safely…"},"finish_reason":null}]}data: {"id":"7c1e…","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}data: {"id":"7c1e…","choices":[],"usage":{"prompt_tokens":19,"completion_tokens":41,"total_tokens":60,"cost":0.003,"cost_source":"estimated"}}data: [DONE]Models
Every model behind the same key and the same balance. Image prices are credits per image; text prices are credits per 1M tokens, input and output.
Image models
Nano Banana API
Google's fast Gemini image model: general-purpose text-to-image and image-to-image at 1K with reference-image input — the lowest price per image in the catalogue.
from $0.025/image · 2 credits
Nano Banana API — pricing and quick start →Nano Banana Pro API
Google's higher-fidelity Gemini image model: more detail and closer prompt adherence, with 1K, 2K and 4K output for hero shots and print-sized crops.
from $0.05/image · 4 credits
Nano Banana Pro API — pricing and quick start →Qwen-Image 2.0 API
Alibaba's Qwen-Image 2.0: strong at rendering legible text inside images, including Chinese, and at illustration styles; 1K and 2K output with reference-image input.
from $0.038/image · 3 credits
Qwen-Image 2.0 API — pricing and quick start →Qwen-Image Max API
The flagship of Alibaba's Qwen-Image line: the highest-fidelity tier for photoreal and text-heavy compositions, with 1K and 2K output and reference-image input.
from $0.10/image · 8 credits
Qwen-Image Max API — pricing and quick start →Seedream 5.0 API
ByteDance's Seedream 5.0: a high-resolution model for photoreal frames and in-image text, served at 2K and 4K, with reference-image input.
from $0.20/image · 16 credits
Seedream 5.0 API — pricing and quick start →Wan 2.7 Image API
Alibaba's Wan 2.7 Image (Tongyi Wanxiang): fast, low-cost text-to-image and reference-image generation at 1K and 2K — for previews and bulk generation.
from $0.025/image · 2 credits
Wan 2.7 Image API — pricing and quick start →Text models
- Qwen3.8-Max API$1.734 in · $5.202 out per 1M tokens
- Claude Opus 5.5 API$1.836 in · $9.18 out per 1M tokens
- Doubao Seed 2.0 Pro API$1.063 in · $5.10 out per 1M tokens
- GPT-6 Astra API$0.85 in · $4.25 out per 1M tokens
- Kimi K3 API$2.463 in · $12.323 out per 1M tokens
- Doubao Seed 2.0 Lite API$0.213 in · $1.063 out per 1M tokens
- Doubao Seed 2.0 Code API$1.063 in · $5.10 out per 1M tokens
- Doubao Seed 2.0 Mini API$0.027 in · $0.85 out per 1M tokens
Simple, credit-based pricing
Buy a credit pack once, spend it as you generate. No subscriptions, no surprise invoices.
How it works
01
Create an API key
Sign up, verify your email, and generate a scoped key from the dashboard.
02
Submit a generation
POST a prompt to /v1/generations for an image, or messages to /v1/chat/completions for a reply — the same key for both.
03
Poll, stream or get webhooked
Poll the image job or wait for the webhook; read chat replies as they stream. Usage and cost come back with every result.
One key. Every image and chat model.
Sign up, create an API key and send the first request in minutes — pay per image or per token, no subscription.