Topaz Labs connector for AI agents

Tools your video enhancement agent reaches for on Topaz Labs, scoped per user.

CALL ANY TOOL

21 tools for video enhancement in Topaz Labs: create and accept video requests, track status and metrics, estimate and track image jobs, and check credits.

topazlabs_accept_video_request

Accept video request

Accept a video request created by topazlabs_create_video_request, which reserves the estimated credits. Returns uploadId and urls (presigned upload URLs) when the source must be uploaded, or a message when the request was created with a source_url and processing has started. Use this after reviewing the estimate. For an upload, PUT the file (or its 500 MB parts, in order) to the returned urls yourself, outside this tool, then call topazlabs_complete_video_upload. Calling it again returns the same URLs. Requires a requestId from topazlabs_create_video_request.

Parameters

Name

Type

Required

Description

request_id

string

Required

ID (UUID) of the video request, returned as requestId by topazlabs_create_video_request or topazlabs_create_express_video_request, or listed by topazlabs_list_video_requests. Example: 0193b36c-b37f-7997-ab75-c85f61b451e8

topazlabs_cancel_image_job

Cancel image job

topazlabs_cancel_video_request

Cancel video request

topazlabs_complete_video_upload

Upload complete video

topazlabs_create_express_video_request

Create express video request

Start a Topaz Labs video enhancement job in one step, without source metadata. Returns requestId, plus uploadId and uploadUrls when no source_url is given, and cost estimates when they can be calculated up front. Use this for the quickest start: with source_url Topaz fetches the video and begins processing at once; without it, PUT the file to uploadUrls[0] yourself and processing starts after the upload. Use topazlabs_create_video_request to see the credit estimate before committing. Poll topazlabs_get_video_request_status for progress and the download link. This spends API credits.

Parameters

Name

Type

Required

Description

container

string

Required

Container format of the source video file, matched case-insensitively. Example: mp4. One of: `3gp`, `avi`, `dv`, `flv`, `m1v`, `m2t`, `m2ts`, `m2v`, `m4v`, `mkv`, `mov`, `mp4`, `mpeg`, `mpg`, `mts`, `mxf`, `ser`, `ts`, `vob`, `webm`, `wmv`.

filters

array

Required

The AI models to run, in order, as a list of filter objects. Each filter needs a model short code. Enhancement models (for example prob-4 Proteus, rhea-1 Rhea, iris-3 Iris, nyx-3 Nyx, slp-2.6 Starlight Precise 2.6, ast-2 Astra 2) upscale, denoise or restore; frame-interpolation models (apo-8 Apollo, chr-2 Chronos, aion-1 Aion) change the frame rate or add slow motion. Call topazlabs_get_video_system_status for the codes the service accepts right now. Enhancement filters also take optional tuning values such as auto (Auto, Relative or Manual), videoType (Progressive or Interlaced), and detail, noise, compression, halo and blur strengths between -1 and 1; frame-interpolation filters take fps (15-240) or slowmo (1-16). Leave the tuning values out to let Topaz pick them. Example: [{"model": "prob-4"}, {"model": "apo-8", "fps": 60}]

output_height

integer

Required

Output video height in pixels. Topaz rounds it up to a multiple of 4; the encoder maximum applies as for output_width. Example: 2160

output_width

integer

Required

Output video width in pixels. Topaz rounds it up to a multiple of 4. The maximum depends on the encoder (4096 for H264, 8192 for H265 and VP9, 16384 for AV1 and ProRes). Without an upscaling model in filters the source width is kept. Example: 3840

audio_bitrate

string

Optional

Audio bitrate in kbps when audio_transfer is Copy or Convert. The codec default is used when empty. Example: 320

audio_codec

string

Optional

Audio codec for the output: AAC, AC3 or PCM. Required when audio_transfer is Copy or Convert. Example: AAC. One of: `AAC`, `AC3`, `PCM`.

audio_transfer

string

Optional

What to do with the audio track: Copy keeps it as is, Convert re-encodes it with audio_codec, None drops it. Copy and Convert also need audio_codec. Defaults to None (no audio) on this endpoint. Example: Copy. One of: `Copy`, `Convert`, `None`.

crop_to_fit

boolean

Optional

Set to true to center-crop the video to the output aspect ratio instead of letterboxing it. Example: false

destination_url

string

Optional

Optional presigned upload (PUT) URL in your own S3-compatible storage. When set, Topaz writes the enhanced video there instead of its own storage. Example: https://my-bucket.s3.amazonaws.com/out/enhanced.mp4?X-Amz-Signature=...

dynamic_compression_level

string

Optional

Automatic quality-based compression: Low, Mid or High. Use either this or video_bitrate. Applies to AV1, H264 and H265, and defaults to High when neither is set. Example: Mid. One of: `Low`, `Mid`, `High`.

md5_hash

string

Optional

Optional MD5 hash of the source file in hex. Topaz uses it to check the file it receives. Example: 5d41402abc4b2a76b9719d911017c592

output_container

string

Optional

Output file container: mp4 (default), mov, mkv, avi or webm. Replaced with mov for ProRes and with mp4 for AV1 and VP9. Example: mp4. One of: `mp4`, `mov`, `mkv`, `avi`, `webm`.

output_frame_rate

number

Optional

Output frame rate in frames per second. Without a frame-interpolation filter Topaz keeps the source frame rate. Needed only when a frame-interpolation filter has no fps of its own. Example: 60

source_url

string

Optional

Presigned download (GET) URL of the source video, for example an S3 presigned URL. Topaz fetches the file from this URL itself and checks that it can read it when the request is created, so no upload is needed. When set, processing starts right away and credits are reserved; no accept or upload step follows. When empty, the response returns uploadUrls and you must HTTP PUT the file to uploadUrls[0] yourself, outside this tool. Example: https://my-bucket.s3.amazonaws.com/in/clip.mp4?X-Amz-Signature=...

video_bitrate

string

Optional

Constant video bitrate with a k (kilobits) or m (megabits per second) suffix. Use either this or dynamic_compression_level, not both. Required for VP9. Capped at 800m for AV1 and 2000m for H264, H265 and VP9. Example: 50m

video_encoder

string

Optional

Video encoder for the output: AV1, H264, H265, ProRes or VP9. Defaults to H265. ProRes always produces a mov file and AV1 and VP9 an mp4 file. VP9 also needs video_bitrate. Example: H265. One of: `AV1`, `H264`, `H265`, `ProRes`, `VP9`.

video_profile

string

Optional

Encoder profile. Accepted values per encoder: AV1 8-bit (default) or 10-bit; H264 High; H265 Main or Main10 (default); ProRes 422 Proxy, 422 LT, 422 Std or 422 HQ (default); VP9 Good (default) or Best. Leave empty for the encoder default. Example: Main10

webhook_url

string

Optional

Optional HTTPS URL that Topaz POSTs a notification to when the request completes or fails. Example: https://example.com/webhooks/topaz

topazlabs_create_video_request

Create video request

Create a Topaz Labs video enhancement request from the source video's metadata and get a cost and time estimate, without spending credits or starting processing. Returns requestId and estimates (cost in credits and time in seconds, each as a [low, high] range). Use this when you want the estimate before committing, then call topazlabs_accept_video_request to reserve credits and start. Use topazlabs_create_express_video_request to start in one step without metadata.

Parameters

Name

Type

Required

Description

audio_transfer

string

Required

What to do with the audio track: Copy keeps it as is, Convert re-encodes it with audio_codec, None drops it. Copy and Convert also need audio_codec. Example: Copy. One of: `Copy`, `Convert`, `None`.

container

string

Required

filters

array

Required

output_frame_rate

number

Required

Output frame rate in frames per second. Without a frame-interpolation filter Topaz keeps the source frame rate. Usually the same as source_frame_rate. Example: 30

output_height

integer

Required

Output video height in pixels. Topaz rounds it up to a multiple of 4; the encoder maximum applies as for output_width. Example: 2160

output_width

integer

Required

source_duration

number

Required

Duration of the source video in seconds. Example: 213

source_frame_count

integer

Required

Total number of frames in the source video. Example: 6371

source_frame_rate

number

Required

Frame rate of the source video in frames per second. Example: 30

source_height

integer

Required

Height of the source video in pixels. Example: 720

source_size

integer

Required

Size of the source video file in bytes; must be under 100 GB. Example: 16125151

source_width

integer

Required

Width of the source video in pixels. Example: 1280

audio_bitrate

string

Optional

Audio bitrate in kbps when audio_transfer is Copy or Convert. The codec default is used when empty. Example: 320

audio_codec

string

Optional

Audio codec for the output: AAC, AC3 or PCM. Required when audio_transfer is Copy or Convert. Example: AAC. One of: `AAC`, `AC3`, `PCM`.

crop_to_fit

boolean

Optional

Set to true to center-crop the video to the output aspect ratio instead of letterboxing it. Example: false

destination_url

string

Optional

dynamic_compression_level

string

Optional

output_container

string

Optional

Output file container: mp4 (default), mov, mkv, avi or webm. Replaced with mov for ProRes and with mp4 for AV1 and VP9. Example: mp4. One of: `mp4`, `mov`, `mkv`, `avi`, `webm`.

source_url

string

Optional

Presigned download (GET) URL of the source video, for example an S3 presigned URL. Topaz fetches the file from this URL itself and checks that it can read it when the request is created, so no upload is needed. When set, topazlabs_accept_video_request starts processing directly; when empty, accept returns upload URLs that you PUT the file to yourself, outside this tool. Example: https://my-bucket.s3.amazonaws.com/in/clip.mp4?X-Amz-Signature=...

video_bitrate

string

Optional

video_encoder

string

Optional

video_profile

string

Optional

webhook_url

string

Optional

Optional HTTPS URL that Topaz POSTs a notification to when the request completes or fails. Example: https://example.com/webhooks/topaz

topazlabs_delete_video_files

Delete video files

topazlabs_estimate_generative_image_job

Estimate generative image job

topazlabs_estimate_image_job

Estimate image job

Estimate the credits and processing time of a Topaz Labs image job on a standard (non-generative) model, without running it. Returns credits and duration (seconds). Use this to price an image enhancement before submitting it. Use topazlabs_estimate_generative_image_job for generative models such as Redefine or Wonder.

Parameters

Name

Type

Required

Description

input_height

integer

Required

Height of the input image in pixels, 1 to 32000. Example: 1080

input_width

integer

Required

Width of the input image in pixels, 1 to 32000. Example: 1920

category

string

Optional

Model category: Enhance (upscaling), Sharpen, Denoise, Restore, Lighting, Matting or Tool. Defaults to Enhance. Example: Enhance. One of: `Enhance`, `Sharpen`, `Denoise`, `Restore`, `Lighting`, `Matting`, `Tool`. Default: `Enhance`.

crop_to_fill

boolean

Optional

Set to true to crop the image to fill a different aspect ratio instead of letterboxing it. Defaults to false. Example: false

model

string

Optional

Standard (non-generative) model name. Defaults to Standard V2. Known values include Standard V2, Low Resolution V2, CGI, High Fidelity V2, Text Refine (Enhance); Standard, Strong, Lens Blur V2, Motion Blur, Natural, Refocus, Wildlife, Portrait (Sharpen); Normal, Strong, Extreme (Denoise); Adjust V2, White Balance, Colorize (Lighting); Object (Matting). Example: High Fidelity V2. Default: `Standard V2`.

model_settings

object

Optional

Optional model-specific settings as string key-value pairs, sent alongside the other fields. Only settings the chosen model understands are used; see the model's page on developer.topazlabs.com. Example: {"faceEnhancement": "true"}

output_format

string

Optional

Output image format: jpeg (default), jpg, png, tiff or tif. Example: png. One of: `jpeg`, `jpg`, `png`, `tiff`, `tif`.

output_height

integer

Optional

Target output height in pixels (Enhance models only). Set one of output_width or output_height and the other scales proportionally. Example: 2160

output_width

integer

Optional

Target output width in pixels (Enhance models only). Set one of output_width or output_height and the other scales proportionally. Example: 3840

topazlabs_estimate_video_cancellation

Estimate video cancellation

topazlabs_get_credit_balance

Get credit balance

topazlabs_get_credit_usage

Get credit usage

topazlabs_get_image_download_url

Get image download url

topazlabs_get_image_input_download_url

Get image input download url

topazlabs_get_image_job_status

Get image job status

topazlabs_get_video_request

Get video request

topazlabs_get_video_request_metrics

Get video request metrics

topazlabs_get_video_request_status

Get video request status

topazlabs_get_video_system_status

Get video system status

topazlabs_list_image_jobs

List image jobs

topazlabs_list_video_requests

List video requests

Build your Agent

Same auth pattern across LangChain, OpenAI, Anthropic, and Google ADK.

import { ScalekitClient } from "@scalekit-sdk/node";
import { createReactAgent } from "@langchain/langgraph/prebuilt";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

// Topaz Labs tools scoped to this user
const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["topazlabs"], toolNames: [
    "topazlabs_create_express_video_request",
    "topazlabs_get_video_request_status",
    "topazlabs_get_credit_balance"] },
  pageSize: 100,
});

const agent = createReactAgent({ llm, tools });
await agent.invoke({ messages: [{ role: "user", content: "Enhance this product demo video and tell me when it is done" }] });

import OpenAI from "openai";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const openai = new OpenAI();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["topazlabs"] }, pageSize: 100,
});

const res = await openai.chat.completions.create({
  model: "gpt-5",
  messages: [{ role: "user", content: "Enhance this product demo video and tell me when it is done" }],
  tools,
});

// Execute the tool call with the user's vaulted Topaz Labs API key
await sk.tools.executeTool(res.choices[0].message.tool_calls[0], "user_123");

import Anthropic from "@anthropic-ai/sdk";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const anthropic = new Anthropic();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["topazlabs"] }, pageSize: 100,
});

const msg = await anthropic.messages.create({
  model: "claude-sonnet-5",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Enhance this product demo video and tell me when it is done" }],
  tools,
});

// Tool call runs with the user's vaulted Topaz Labs API key
await sk.tools.executeTool(msg.content, "user_123");

import { Agent } from "@google/adk/agents";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["topazlabs"] }, pageSize: 100,
});

const agent = new Agent({
  name: "video_enhancement_agent",
  model: "gemini-2.5-pro",
  instruction: "Topaz Labs tools scoped to this user",
  tools,
});

await agent.run("Enhance this product demo video and tell me when it is done");

Try these prompts

Copy any prompt into your agent. Each maps directly to a Topaz Labs tool. Click to copy, paste into your agent, done.

Video

Copy the prompt

Copied

Enhance the video at this URL with an express request.

Copy the prompt

Copied

How far along is the product demo video?

Copy the prompt

Copied

Cancel the webinar video request and show the refund.

Images

Copy the prompt

Copied

Estimate the credits to enhance a 12 MP photo with Standard V2.

Copy the prompt

Copied

Get the download link for my finished image job.

Copy the prompt

Copied

List my image jobs from today.

Credits and service

Copy the prompt

Copied

What is our Topaz Labs credit balance?

Copy the prompt

Copied

Show credit usage by day for the last week.

Copy the prompt

Copied

Is the Topaz Labs video service available right now?

SEE HOW AUTH WORKS

Each user connects their own Topaz Labs API key once, and Scalekit sends it with every call. Keys stay vaulted, every call is scope checked, and every action is logged.

Authorize

Your user connects

Topaz Labs

once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access

Who:

user ‘A’

when:

Once per user

access:

Limited to user

Store

Their

Topaz Labs

token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection

vault:

encrypted

scope:

per-user

tokens:

auto-refreshed

Resolve

When your agent calls a

Topaz Labs

tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs

speed:

~40ms

check:

before every call

seen by:

nobody

Audit

Every

Topaz Labs

tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it

history:

90 days

export:

SIEM-ready

logged:

every call

Test other agents

See the same per-user auth pattern across other media and AI connectors.

GTM and RevOps Teams

Competitive intelligence briefing agent

Scans Gong calls for competitor mentions, matches each one to its Notion battlecard, and DMs every affected rep a single Slack digest per cycle. Every call runs as the PMM who owns the briefing, never a shared bot.

Why Scalekit

Secure your agent's access. Connectors ship in minutes

01.

Shared tokens break per-user analytics

A shared Topaz Labs API key looks fine in a demo. In production every video request and credit spend looks like one service account, and you cannot tell which user triggered it. Scalekit resolves the credential of the actual user who triggered the agent, never a shared bot.

// shared key
audit → bot_service_account

// scalekit
audit → user_abc ✓

02.

Authentication is not authorization

03.

Multi-tenancy is architectural

04.

Topaz Labs today. Ten connectors tomorrow.

“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”

Venu Madhav Kattagoni

Head of Engineering / Von

FAQs

Frequently Asked Questions

Does the agent access Topaz Labs as the user or through a shared key?

As the user. Each user connects their own Topaz Labs API key once, and Scalekit sends it with every call. Audit logs attribute every video request and credit spend to that user, not a shared service account.

Where is the Topaz Labs API key stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. To rotate it, the user reconnects with a new API key. Revocation is a single dashboard action. Credentials never appear in prompts, logs, or LLM context.

Can I limit what the agent does in Topaz Labs?

Yes. Filter by tool name in listScopedTools to expose only what you want, for example the status and estimate tools without topazlabs_accept_video_request, which reserves credits and starts processing, or topazlabs_delete_video_files, which permanently deletes the source and enhanced files. Scalekit also enforces scope checks before every API call.

What happens when a user revokes Topaz Labs access?

The connection is invalidated on the next tool call. Subsequent requests for that user fail closed with a clear error. Other users in the tenant remain unaffected. The event is logged for audit.

Can the agent submit image jobs to Topaz Labs?

No. The image tools estimate credits and processing time, list jobs, track status, cancel pending jobs, and return download links for jobs your own application submits to the Topaz Image API. Video enhancement requests can be created and started from the agent.