PiAPI (Windsurf)
POST /api/v1/task → submit
GET /api/v1/task/{id} → poll
DELETE /api/v1/task/{id} → cancel (only while pending)
POST /v1/chat/completions → LLM (OpenAI-compatible, supports SSE)
Auth: x-api-key: <PIAPI_API_KEY> (lowercase header).
Setup
export PIAPI_API_KEY="pk-..."
export PIAPI_WEBHOOK_SECRET="..." # optional, webhook verify only
piapi-cli --help
No official PiAPI SDK exists. The CLI uses requests.
Cheatsheet
piapi-cli submit --model <m> --task-type <t> --input '<json>'
piapi-cli result <task_id>
piapi-cli wait <task_id> --timeout 1800 --poll-interval 2
piapi-cli cancel <task_id>
piapi-cli run --model <m> --task-type <t> --input '<json>'
piapi-cli imagine "prompt" --aspect-ratio 16:9 --process-mode fast --wait
piapi-cli flux --prompt "..." --model Qubico/flux1-schnell --wait
piapi-cli kling --image-url URL --prompt "..." --duration 5 --wait
piapi-cli suno --prompt "lofi instrumental" --wait
piapi-cli faceswap --target-url T --swap-url F --wait
piapi-cli llm --model gpt-4o-mini \
--message 'system:You are concise.' --message 'user:Hello.'
piapi-cli verify-webhook --header-secret "$RAW_HEADER" --expected "$PIAPI_WEBHOOK_SECRET"
Status enum drift
Lowercase before comparing. Terminal-success: completed | complete | success | succeeded. Terminal-failure: failed | failure | error | canceled | cancelled | rejected.
Capitalized family (Midjourney, Kling, Faceswap):
Pending | Processing | Staged | Completed | Failed.
Lowercase family (Flux, Veo 3, Gemini, Suno):
pending | starting | processing | success | failed | retry.
Common model · task_type
| Family |
model |
task_type |
| Midjourney |
midjourney |
imagine, upscale, variation, inpaint |
| Flux |
Qubico/flux1-schnell / Qubico/flux1-dev |
txt2img, img2img, inpaint, controlnet-lora |
| Kling |
kling |
text2video, image2video, extend |
| Luma |
luma |
text2video, image2video, extend |
| Hailuo |
hailuo |
text2video, image2video, subject2video |
| Veo 3 |
veo3 |
txt2vid, img2vid |
| Seedance |
seedance |
text-to-video, image-to-video |
| Suno |
music-u |
generate_music, generate_music_custom, extend |
| Hunyuan |
Qubico/hunyuan |
txt2video-lora, img2video-lora |
| MMAudio |
Qubico/mmaudio |
video2audio |
| Faceswap |
Qubico/image-toolkit / Qubico/video-toolkit |
face-swap, multi-face-swap |
| Trellis |
Qubico/trellis |
image-to-3d |
| F5-TTS |
Qubico/f5-tts |
txt2speech |
| Gemini Image |
gemini |
nano-banana-text-to-image, nano-banana-edit |
Refs
1---2name: piapi-73description: PiAPI inference gateway. One async-task envelope plus an OpenAI-compatible LLM proxy. Aggregates Midjourney, Flux, Kling, Luma, Hailuo, Veo 3, Suno, Hunyuan, MMAudio, F5-TTS, Trellis, Gemini Nano Banana, Seedance, and Faceswap.4---56# PiAPI (Windsurf)78```9POST /api/v1/task → submit10GET /api/v1/task/{id} → poll11DELETE /api/v1/task/{id} → cancel (only while pending)12POST /v1/chat/completions → LLM (OpenAI-compatible, supports SSE)13```1415Auth: `x-api-key: <PIAPI_API_KEY>` (lowercase header).1617## Setup1819```bash20export PIAPI_API_KEY="pk-..."21export PIAPI_WEBHOOK_SECRET="..." # optional, webhook verify only22piapi-cli --help23```2425No official PiAPI SDK exists. The CLI uses `requests`.2627## Cheatsheet2829```bash30piapi-cli submit --model <m> --task-type <t> --input '<json>'31piapi-cli result <task_id>32piapi-cli wait <task_id> --timeout 1800 --poll-interval 233piapi-cli cancel <task_id>34piapi-cli run --model <m> --task-type <t> --input '<json>'3536piapi-cli imagine "prompt" --aspect-ratio 16:9 --process-mode fast --wait37piapi-cli flux --prompt "..." --model Qubico/flux1-schnell --wait38piapi-cli kling --image-url URL --prompt "..." --duration 5 --wait39piapi-cli suno --prompt "lofi instrumental" --wait40piapi-cli faceswap --target-url T --swap-url F --wait4142piapi-cli llm --model gpt-4o-mini \43 --message 'system:You are concise.' --message 'user:Hello.'4445piapi-cli verify-webhook --header-secret "$RAW_HEADER" --expected "$PIAPI_WEBHOOK_SECRET"46```4748## Status enum drift4950Lowercase before comparing. Terminal-success: `completed | complete | success51| succeeded`. Terminal-failure: `failed | failure | error | canceled |52cancelled | rejected`.5354Capitalized family (Midjourney, Kling, Faceswap):55`Pending | Processing | Staged | Completed | Failed`.56Lowercase family (Flux, Veo 3, Gemini, Suno):57`pending | starting | processing | success | failed | retry`.5859## Common model · task_type6061| Family | model | task_type |62|---|---|---|63| Midjourney | `midjourney` | `imagine`, `upscale`, `variation`, `inpaint` |64| Flux | `Qubico/flux1-schnell` / `Qubico/flux1-dev` | `txt2img`, `img2img`, `inpaint`, `controlnet-lora` |65| Kling | `kling` | `text2video`, `image2video`, `extend` |66| Luma | `luma` | `text2video`, `image2video`, `extend` |67| Hailuo | `hailuo` | `text2video`, `image2video`, `subject2video` |68| Veo 3 | `veo3` | `txt2vid`, `img2vid` |69| Seedance | `seedance` | `text-to-video`, `image-to-video` |70| Suno | `music-u` | `generate_music`, `generate_music_custom`, `extend` |71| Hunyuan | `Qubico/hunyuan` | `txt2video-lora`, `img2video-lora` |72| MMAudio | `Qubico/mmaudio` | `video2audio` |73| Faceswap | `Qubico/image-toolkit` / `Qubico/video-toolkit` | `face-swap`, `multi-face-swap` |74| Trellis | `Qubico/trellis` | `image-to-3d` |75| F5-TTS | `Qubico/f5-tts` | `txt2speech` |76| Gemini Image | `gemini` | `nano-banana-text-to-image`, `nano-banana-edit` |7778## Refs7980- https://piapi.ai/docs/overview81- https://piapi.ai/docs/llm-api82- https://piapi.ai/pricing
wesleysimplicio