Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
// SPDX-License-Identifier: AGPL-3.0-only
// Copyright 2026-present the Unsloth AI Inc. team. All rights reserved. See /studio/LICENSE.AGPL-3.0
// Which tool panels a model gets. Free of JSX so the node test runner can load it.
import type { AudioConvertCaps } from "@/features/chat/types/api";
import type { AudioOptionSpec } from "../audio-options";
import {
type NativeAudioInstructionsKind,
nativeAudioInstructionsKind,
} from "../audio-page-policy";
import { isMusicGenerationModel } from "../catalog";
import { parseMusicCapabilities } from "../music/music-types";
import type { AudioWorkflowId } from "../workflows";
import type {
AnyAudioToolPanel,
AudioModelContext,
AudioReferenceTextMode,
AudioRunPatch,
AudioToolPanel,
CoreInputs,
} from "./types";
export function instructionsKindFor(
ctx: AudioModelContext,
): NativeAudioInstructionsKind | null {
return ctx.musicGeneration
? "music"
: nativeAudioInstructionsKind(ctx.audioType);
}
/** Only a loaded music model without studio modes (native MiniMax) keeps the old description. */
export function legacyMusicDescription(ctx: AudioModelContext): boolean {
return ctx.audioMusic !== true && isMusicGenerationModel(null, ctx.audioType);
}
export function panelApplies(
panel: Pick<AudioToolPanel<unknown>, "workflows" | "families" | "appliesTo">,
workflow: AudioWorkflowId,
ctx: AudioModelContext,
): boolean {
if (!panel.workflows.includes(workflow)) return false;
return panel.appliesTo
? panel.appliesTo(ctx)
: panel.families.includes(ctx.audioFamily ?? "");
}
export function claimedOptionNames(
panels: readonly Pick<AudioToolPanel<unknown>, "claims">[],
): Set<string> {
return new Set(panels.flatMap((panel) => panel.claims));
}
const REFERENCE_TEXT_MODES: ReadonlySet<string> = new Set([
"required",
"optional",
"unused",
]);
export function audioModelContextFor(
status: {
audio_type?: string | null;
audio_family?: string | null;
audio_workflows?: readonly string[] | null;
audio_required_inputs?: readonly string[] | null;
audio_request_inputs?: readonly string[] | null;
audio_reference_text?: string | null;
audio_convert?: AudioConvertCaps | null;
audio_music?: unknown;
audio_clone?: unknown;
} | null,
page: {
musicGeneration: boolean;
cudaMusicGeneration: boolean;
musicNeedsDescription: boolean;
},
): AudioModelContext {
const referenceText = status?.audio_reference_text;
const clone = status?.audio_clone;
return {
audioType: status?.audio_type ?? null,
audioFamily: status?.audio_family ?? null,
musicGeneration: page.musicGeneration,
cudaMusicGeneration: page.cudaMusicGeneration,
musicNeedsDescription: page.musicNeedsDescription,
audioWorkflows: Array.isArray(status?.audio_workflows)
? status.audio_workflows
: [],
requiredInputs: Array.isArray(status?.audio_required_inputs)
? status.audio_required_inputs
: [],
referenceTextMode:
typeof referenceText === "string" &&
REFERENCE_TEXT_MODES.has(referenceText)
? (referenceText as AudioReferenceTextMode)
: null,
requestInputs: Array.isArray(status?.audio_request_inputs)
? status.audio_request_inputs
: null,
audioMusic: parseMusicCapabilities(status?.audio_music) !== null,
convert: convertCapsOf(status?.audio_convert),
cloneDescription: isPlainObject(clone) && clone.description === true,
};
}
/** Whether the model reads one of these request inputs; unknown (null) counts as yes. */
export function readsRequestInput(
ctx: Pick<AudioModelContext, "requestInputs">,
...names: string[]
): boolean {
return (
ctx.requestInputs == null ||
names.some((name) => ctx.requestInputs?.includes(name))
);
}
function convertCapsOf(value: unknown): AudioConvertCaps | null {
if (!isPlainObject(value)) return null;
const caps = value as Partial<AudioConvertCaps>;
const modes = Array.isArray(caps.modes)
? caps.modes.filter((mode) => mode === "speech" || mode === "singing")
: [];
if (modes.length === 0) return null;
return {
modes,
target: caps.target === "builtin" ? "builtin" : "audio",
builtin_voices: Array.isArray(caps.builtin_voices)
? caps.builtin_voices.filter(
(voice) =>
isPlainObject(voice) &&
typeof voice.id === "string" &&
typeof voice.label === "string",
)
: [],
pitch: isPlainObject(caps.pitch) ? caps.pitch : {},
style: caps.style === true,
route_reloads: caps.route_reloads === true,
source_max_seconds:
typeof caps.source_max_seconds === "number" && caps.source_max_seconds > 0
? caps.source_max_seconds
: 300,
};
}
export function toolValueKey(
model: string | null | undefined,
workflow: AudioWorkflowId,
panelId: string,
): string {
return `${model ?? ""}:${workflow}:${panelId}`;
}
const isPlainObject = (value: unknown): value is Record<string, unknown> =>
typeof value === "object" && value !== null && !Array.isArray(value);
// Kept value over defaults, so a value saved by an older build still has every field.
export function panelValue<V>(
panel: Pick<AudioToolPanel<V>, "id" | "initial">,
values: Readonly<Record<string, unknown>>,
specs: AudioOptionSpec[],
): V {
const initial = panel.initial(specs);
const stored = values[panel.id];
if (stored === undefined) return initial;
if (isPlainObject(initial) && isPlainObject(stored)) {
return { ...initial, ...stored } as V;
}
return stored as V;
}
export function collectToolRequest(
panels: readonly AnyAudioToolPanel[],
values: Readonly<Record<string, unknown>>,
core: CoreInputs,
ctx: AudioModelContext,
specs: AudioOptionSpec[] = [],
): { patch: AudioRunPatch; error: string | null } {
const patch: AudioRunPatch = {};
let error: string | null = null;
for (const panel of panels) {
const value = panelValue(panel, values, specs);
error ??= panel.validate?.(value, core, ctx) ?? null;
const part = panel.toRequest(value, ctx);
if (part.options) patch.options = { ...patch.options, ...part.options };
if (part.inputs) patch.inputs = { ...patch.inputs, ...part.inputs };
if (part.convert) patch.convert = { ...patch.convert, ...part.convert };
for (const key of [
"instructions",
"language",
"route",
"text",
"speed",
"referenceTextMode",
] as const) {
if (part[key] !== undefined) {
(patch as Record<string, unknown>)[key] = part[key];
}
}
}
return { patch, error };
}