Parallel Alternatives

Framing

The problem

Serial iteration wastes turns when users are still searching the space.

The pattern

Generate distinct candidates simultaneously for comparison and blending.

Why chat breaks here

One-answer-at-a-time chat narrows exploration too early.

Risks

Too many alternatives can create evaluation fatigue.

Avoid when

A single high-confidence answer is preferable to divergent exploration.

Use when

Users are still searching the solution space and one answer narrows it too early.

DOPE evaluation

Directability

Select, blend, or regenerate any variant independently

Observability

All alternatives visible simultaneously — no serial scrolling

Predictability

Each alternative has a labeled approach, tone, and length

Explainability

Each alternative's tone, approach, and length rationale are labeled

In the wild

• Midjourney Grid (Midjourney) — Every job returns four images at once. On web the grid arrives pre-separated, and hovering any image offers Vary (Subtle) or Vary (Strong), each returning a fresh grid of four. Still the canonical example — though the V1–V4 buttons are the Discord path, now filed under Legacy Features.
• Suno Song Versions (Suno) — Every Create returns two complete takes of the same prompt — different arrangement, different vocal performance — and you keep, discard or extend either one. Still two per run on the current v5.5 model; the V4 / V4.5 / V4.5+ wording predates two model generations. Audio-domain proof the pattern travels.
• Adobe Firefly Variants (Adobe) — Adobe's own Image 4 and Image 4 Ultra still return four variants per prompt, and the Vary control offers More like this, Keep style or Keep subject on any result. The count is per-model now, though: Image 5 and the partner models (Imagen, Flux, GPT Image) return a single image.
• Google Stitch (Google) — Voice prompts like "show me three menu options" produce parallel design alternatives directly on the same infinite canvas — variants as a first-class canvas operation.
• Runway Gen-4 Image (Runway) — Gen-4 Image ships a Quantity control — one image or a batch of four from a single prompt, with references steering all four. The video side does not do this: Gen-4.5 returns one clip per run and you compare across a scrollable Session instead, which is serial iteration wearing a feed.
• Cursor /best-of-n (Anysphere) — Runs one task across several models at once, each in its own git worktree so the candidates stay isolated from each other and from your checkout. Cursor is explicit that best-of-n compares runs only — nothing merges until you pick a winner and commit from that worktree or run /apply-worktree.

FAQ

When should I use the Parallel Alternatives pattern?

Users are still searching the solution space and one answer narrows it too early.

When should I avoid the Parallel Alternatives pattern?

A single high-confidence answer is preferable to divergent exploration.

What problem does Parallel Alternatives solve?

Serial iteration wastes turns when users are still searching the space.

Why is chat the wrong fit for this?

One-answer-at-a-time chat narrows exploration too early.

Related patterns

• Often paired with: Region Lock — Lock the parts that work, only let alternatives diverge in the unlocked rest.
• Often paired with: Select & Transform — Transform a selection into three candidate edits in parallel.
• Alternative to: Tone of Voice — Three rendered candidates vs one direct dial for stylistic exploration.