- Category: media generation
- Auth: API_KEY
- Composio-managed OAuth available? N/A
- Tools: 13
- Triggers: 0
- Slug:
FAL_AI - Version: 20261101_00
- Instant: Supported on the latest version. Support varies by tool.
Tools
Cancel Queue Request
Cancel Queue Request
Slug:
FAL_AI_CANCEL_QUEUE_REQUESTRequest cancellation of a queued or in-progress fal.ai job using your own fal account. Queued jobs are removed; for jobs already running, fal sends a cancellation signal and the model determines how to stop.Input Parameters
The model identifier in namespace/name format (e.g., ‘fal-ai/flux’, ‘fal-ai/fast-sdxl’)
Output
Estimate Pricing
Estimate Pricing
Slug:
FAL_AI_ESTIMATE_PRICINGEstimate expected costs for API calls or unit-based usage across one or more fal.ai model endpoints using your own fal account.Input Parameters
Map of endpoint IDs (e.g., ‘fal-ai/flux/dev’) to their quantity specifications. Use call_quantity for historical_api_price or unit_quantity for unit_price
Estimation method: ‘historical_api_price’ for call-based estimates or ‘unit_price’ for unit-based estimates
Output
Get JWKS for Webhook Verification
Get JWKS for Webhook Verification
Get Models
Get Models
Slug:
FAL_AI_GET_MODELSDiscover and search fal.ai model endpoints using your own fal account. Find models by ID, category or query, with pagination and optional OpenAPI schema expansion.Input Parameters
Free-text search query. Filters models by name, description, or category. Use for broad text-based searches.
Fields to expand in response. Supports ‘openapi-3.0’ to include full OpenAPI schema for each model.
Filter models by status. Use ‘active’ for currently supported models or ‘deprecated’ for outdated ones. Omit to include all statuses.
Filter models by specific category. Use exact category names like ‘text-to-image’, ‘image-to-video’, etc.
Specific endpoint ID(s) to retrieve. Accepts a single string or array of strings (1-50 models). Use to filter models by their unique identifiers.
Output
Get Model Pricing
Get Model Pricing
Slug:
FAL_AI_GET_PRICINGRetrieve unit pricing for specific fal.ai model endpoints using your own fal account, including unit price, billing unit and currency.Input Parameters
Filter by specific endpoint ID(s). Accepts 1-50 endpoint IDs. Can be a single endpoint ID string or a list of endpoint IDs. Examples: ‘fal-ai/flux/dev’ or [‘fal-ai/flux/dev’, ‘fal-ai/flux/pro’]
Output
Get Queue Request Result
Get Queue Request Result
Slug:
FAL_AI_GET_QUEUE_REQUEST_RESULTTool to retrieve the final result of a completed queue request. Use when you need to get the output of a model request that was submitted to the queue and has finished processing. Checks the job once without blocking and returns its pending or failed status or the completed output. With your own fal account it returns the completed output. Free; no extra charge.Instant: Supported on the latest version.Instant pricing: $0 per requestInput Parameters
Use exactly the same model_id you supplied to Submit Async Job. Do not shorten the model path.
Copy the complete request_id returned by Submit Async Job unchanged, using the same user and project. Without your own fal account this is a signed ID with a model prefix; do not replace it with a raw fal request UUID. With your own fal account, use the fal request UUID.
Output
List Model Requests
List Model Requests
Check Queue Request Status
Check Queue Request Status
Stream Request Status Updates
Stream Request Status Updates
Run Model (Sync)
Run Model (Sync)
Slug:
FAL_AI_RUN_MODEL_SYNCRun a fal.ai model synchronously using your own fal account and return its model-specific output. The call performs paid inference and waits for the response.Input Parameters
Model-specific input payload posted as the JSON request body. Fields depend on the chosen model — for example image generation models typically expect a ‘prompt’ (and optionally ‘image_size’, ‘num_inference_steps’, etc.). Refer to the model’s OpenAPI schema (see GET_MODELS with expand=openapi-3.0) for the full field list.
The model identifier in namespace/name format (e.g., ‘fal-ai/flux/schnell’, ‘fal-ai/fast-sdxl’). Slashes in the model id are preserved in the URL path.
Optional session affinity hint to route subsequent requests to the same runner. Sent as the ‘X-Fal-Runner-Hint’ header.
Optional server-side deadline in seconds to start processing, including queue wait (does not limit running inference). Sent as the ‘X-Fal-Request-Start-Timeout’ header.
Output
Submit Async Inference Job
Submit Async Inference Job
Slug:
FAL_AI_SUBMIT_ASYNC_JOBStart image, text-to-video, image-to-video or 3D generation and return a job ID immediately. Without your own fal account, choose one of the 19 model_id/input examples, omit the five queue options that need your own fal account (webhook_url, priority, timeout_seconds, runner_hint, no_retry), then save request_id and use GET_QUEUE_REQUEST_RESULT with the same model_id, user and complete ID to check progress and retrieve the output. The exact Sunburst route supports text-to-image only; Flare, GPT Image 2 and image editing are not substitutes. Status/result retrieval is included in the accepted-job price. Without your own fal account, automatic retries are disabled; do not submit a duplicate after an uncertain response. Other queue helpers, cancellation, streams and webhooks are available only with your own fal account.Instant: Supported on the latest version.Instant pricing: Images from $0.003; videos from $0.04 per second or $0.19 per clip; $2 max per job.Input Parameters
Model settings. Use a complete example for your model and change the prompt or image_url. Schnell: set image_size={width:512,height:512}, num_images=1, num_inference_steps=1..4, enable_safety_checker=true and output_format=png or jpeg. Wan text-to-video and image-to-video: choose resolution=480p/580p/720p and num_frames=17/33/49/65/81; explicitly set frames_per_second=16, aspect_ratio=16:9, num_inference_steps=27, both safety checks=true, enable_prompt_expansion=false, interpolator_model=none and num_interpolated_frames=0. Image-to-video also requires a publicly accessible HTTPS image_url. Trellis requires image_url, ss_guidance_strength=7.5, ss_sampling_steps=12, slat_guidance_strength=3, slat_sampling_steps=12, mesh_simplify=0.95 and texture_size=1024. For the added models, copy the complete paired example: MiniMax requires its prompt optimizer setting and duration where offered; Nano Banana requires one image, resolution and its fixed safety/add-on settings; Kling requires duration and safety settings; Kling 3 also requires start_image_url and generate_audio; Trellis 2 requires image_url and resolution. Seedream Lite requires auto_2K, num_images=1, max_images=1, safety enabled and sync/BytePlus URLs disabled; Pro requires one 1024x1024 or 2048x2048 image, jpeg/png, standard prompt enhancement, safety enabled and sync disabled. GPT Image 2.5 Sunburst requires exact 1024x1024 size, opaque background, low quality, one PNG, sync_mode=false and safety enabled; editing, compression and other fields are rejected. Standard Seedance 2.0 text-to-video requires duration=4, resolution=720p, aspect_ratio=16:9, bitrate_mode=standard and codec=auto; generate_audio explicitly selects native audio at no price difference. Its documented schema has no safety option, so omit safety fields. Without your own fal account these settings are mandatory and unlisted fields or paid add-ons are rejected. With your own fal account, any of the selected model’s input fields are accepted.
Without your own fal account, only the 19 model IDs are accepted listed in the complete paired examples: Schnell, Wan, Trellis, MiniMax Image-01/Hailuo 2.3, Nano Banana 2, Kling 2.5 Turbo Pro, Kling 3 Pro and Trellis 2, Seedream 5.0 Lite and Pro, plus exact GPT Image 2.5 Sunburst text-to-image and standard Seedance 2.0 text-to-video. Flare, GPT Image 2, Sunburst editing and standalone TTS/audio models are not admitted; Kling 3 and Seedance 2.0 native video audio are supported. With your own fal account, any model identifier in namespace/name format is accepted (e.g., ‘fal-ai/flux/schnell’, ‘fal-ai/fast-sdxl’). Slashes in the model id are preserved in the URL path.
Only available with your own fal account; otherwise omit it (retries are disabled automatically). Set to true to disable fal.ai’s automatic retries on transient runner errors. Sent as the ‘X-Fal-No-Retry’ header when true.
Only available with your own fal account; otherwise omit it. Optional queue priority hint. Sent as the ‘X-Fal-Queue-Priority’ header.
Only available with your own fal account; otherwise omit it. Optional session affinity hint to route subsequent requests to the same runner. Sent as the ‘X-Fal-Runner-Hint’ header.
Only available with your own fal account; otherwise omit it. Optional URL to receive a POST callback when the job completes, instead of (or in addition to) polling. Sent to fal.ai as the ‘fal_webhook’ query parameter.
Only available with your own fal account; otherwise omit it. Maximum seconds to wait for processing to start, including queue wait; does not limit running inference. Sent as X-Fal-Request-Start-Timeout.
Output
Subscribe to Async Inference Job
Subscribe to Async Inference Job
Slug:
FAL_AI_SUBSCRIBE_ASYNC_JOBSubmit a fal.ai inference job using your own fal account, poll until it completes or the wait deadline is reached, and return the final result. A timeout does not cancel the underlying job.Input Parameters
The model identifier in namespace/name format (e.g., ‘fal-ai/flux/schnell’). Slashes are preserved in the URL path.
Optional queue priority hint for the initial submission (‘X-Fal-Queue-Priority’).
Optional server-side deadline in seconds to start processing, including queue wait (does not limit running inference) for the initial submission (‘X-Fal-Request-Start-Timeout’).
Hard deadline. If the job hasn’t transitioned to COMPLETED within this many seconds the action raises a timeout error (the underlying job is NOT cancelled). Capped at 900 to keep the action well below Composio’s serverless execution limits.
Output
Upload Input File
Upload Input File
Slug:
FAL_AI_UPLOAD_FILEUpload an input media file to fal.ai storage using your own fal account and return its access URL for use as a model input.Input Parameters
File to upload to fal.ai storage. The returned access_url can then be passed as a model input field (e.g., ‘image_url’, ‘audio_url’, ‘video_url’). Supported types include images (jpg, jpeg, png, gif, webp), videos (mp4, mov, webm), and audio (mp3, wav, flac).