[Partner Nodes] feat(ByteDance): add support for Seed Audio 1.0

[bigcat88]

PR adds one universal node that generates speech, background music, sound effects, and multi-speaker dialogue from a single prompt with ByteDance Seed Audio 1.0

API Node PR Checklist

Scope

• Is API Node Change

Pricing & Billing

• Need pricing update
• No pricing update

If Need pricing update:

• Metronome rate cards updated
• Auto‑billing tests updated and passing

QA

• QA done
• QA not required

Comms

• Informed Kosinkadink

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR adds a new ByteDanceSeedAudioNode to comfy_api_nodes/nodes_bytedance.py for ByteDance Seed Audio 1.0 text-to-speech. It introduces audio, image, and preset-voice reference modes, prompt and connection validation helpers, request construction and response decoding, and node registration. It also adds upscale_image_tensor_to_min_pixels in comfy_api_nodes/util/conversions.py and re-exports it from comfy_api_nodes/util/__init__.py.

Changes

Related issues: None identified.
Related PRs: None identified.
Suggested labels: enhancement, API nodes
Suggested reviewers: maintainers familiar with API node integrations

Sequence Diagram(s)

Included in the hidden review stack artifact above.

Poem
A rabbit hums a byte-sized tune, 🐰
Text turned to voice beneath the moon.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly and concisely summarizes the main change: adding Seed Audio 1.0 support to ByteDance nodes.
Description check	✅ Passed	The description is directly related to the PR and accurately describes the new Seed Audio 1.0 node.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

comfy_api_nodes/nodes_bytedance.py (1)

2751-2756: 🚀 Performance & Scalability | 🔵 Trivial | ⚡ Quick win

Offload the new audio/image conversions from execute.

MP3 encoding, image upscaling/PNG serialization, and response audio decoding can block the event loop for large clips. Move these conversions to a background executor.

Based on learnings, new synchronous CPU/IO-heavy helpers added inside async def execute in comfy_api_nodes/ should be offloaded with asyncio.to_thread or an equivalent executor.

Suggested offload shape

+import asyncio
 import base64
@@
-                mp3_bytes = audio_input_to_mp3(clip).getvalue()
+                mp3_bytes = await asyncio.to_thread(lambda: audio_input_to_mp3(clip).getvalue())
                 references.append(SeedAudioReference(audio_data=base64.b64encode(mp3_bytes).decode("utf-8")))
         elif mode == MODE_IMAGE:
-            image = upscale_image_tensor_to_min_pixels(image, 160_000)
-            references = [SeedAudioReference(image_data=tensor_to_base64_string(image, mime_type="image/png"))]
+            image = await asyncio.to_thread(upscale_image_tensor_to_min_pixels, image, 160_000)
+            image_data = await asyncio.to_thread(tensor_to_base64_string, image, mime_type="image/png")
+            references = [SeedAudioReference(image_data=image_data)]
@@
-        return IO.NodeOutput(audio_bytes_to_audio_input(base64.b64decode(response.audio)))
+        output_audio = await asyncio.to_thread(lambda: audio_bytes_to_audio_input(base64.b64decode(response.audio)))
+        return IO.NodeOutput(output_audio)

Also applies to: 2779-2779

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@comfy_api_nodes/nodes_bytedance.py` around lines 2751 - 2756, The new
synchronous media conversions inside execute are blocking the event loop;
offload the CPU/IO-heavy audio and image work to a background thread. Update the
audio branch in execute so validate_audio_duration/audio_input_to_mp3/base64
encoding runs via asyncio.to_thread or an equivalent executor, and do the same
for image upscaling and tensor_to_base64_string in the MODE_IMAGE path. Also
apply the same offloading pattern to the response audio decoding helper
referenced in the other flagged location so execute stays non-blocking.

Source: Learnings

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@comfy_api_nodes/nodes_bytedance.py`:
- Around line 2541-2544: The prompt tag validation around max_audio_tag() and
the preset voice-mode path is too permissive because `@Audio0` is treated the same
as “no tags,” and `@Audio1` can slip through even when tags should be unused.
Tighten the validation in the prompt/API preparation flow so any explicit
`@AudioN` is rejected when N is invalid or unsupported for the selected mode, and
make the check distinguish “no tag present” from “tag value 0”; update the
relevant validation in max_audio_tag() and the code that consumes it around the
audio request setup to fail before the API call.

---

Nitpick comments:
In `@comfy_api_nodes/nodes_bytedance.py`:
- Around line 2751-2756: The new synchronous media conversions inside execute
are blocking the event loop; offload the CPU/IO-heavy audio and image work to a
background thread. Update the audio branch in execute so
validate_audio_duration/audio_input_to_mp3/base64 encoding runs via
asyncio.to_thread or an equivalent executor, and do the same for image upscaling
and tensor_to_base64_string in the MODE_IMAGE path. Also apply the same
offloading pattern to the response audio decoding helper referenced in the other
flagged location so execute stays non-blocking.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: f8ed4574-713a-4806-94fd-240b5e19ef01

📥 Commits

Reviewing files that changed from the base of the PR and between 35c1470 and 63fcb85.

⛔ Files ignored due to path filters (1)

• comfy_api_nodes/apis/bytedance.py is excluded by !comfy_api_nodes/apis/**

📒 Files selected for processing (3)

• comfy_api_nodes/nodes_bytedance.py
• comfy_api_nodes/util/__init__.py
• comfy_api_nodes/util/conversions.py