Ultimate Guide to Small-Batch Rapid Prototyping

You need 10–1000 parts fast, on budget, and good enough to ship or validate. This guide gives you a practical way to choose the process that minimizes total landed cost per accepted part at those quantities—without drowning in one-by-one process comparisons. Use the one-page matrix, sanity-check with the benchmarks, and follow the example paths to get from prototype to a reliable bridge run.

Key takeaways

• Optimize for total landed cost per accepted part, not just the quote price. Include NRE or tooling, finishing, yield, shipping, and time-to-market effects.

• For 10–50 units, SLS or SLA often wins on speed and cost. CNC is the right call when tolerances or metals drive the spec.

• For 50–300 units, vacuum casting gets powerful for production-like plastics and cosmetics. SLS still excels for functional nylon parts.

• By 200–1000 units, rapid aluminum tooling tends to amortize best when you need repeatability and production resins.

• Tight tolerances, high cosmetics, undercuts, and special certifications can flip the decision.

• Mix methods across phases. For many programs, the lowest-cost path is a staged plan, not a single process from day one.

  
  

Small-batch rapid prototyping decision matrix

Quantity is your primary axis; the strongest secondary selector is the tightest constraint you can’t relax today. Start here, then read the adjustment notes.

Adjustment notes

• If the part has deep undercuts or no draft and must be molded, expect higher NRE for slides and lifters; that may keep SLS or vacuum casting competitive longer.

• If you need SPI-A or tight color matching, move SLA earlier for master quality, or lean into vacuum casting or molding for stable cosmetics.

• If your tightest tolerance is under ±0.1 mm on multiple features, CNC or molding with robust process controls is usually required.

How to read the matrix

Total landed cost per accepted part includes more than the unit quote. Add up the non-recurring engineering or tooling, expected rework and scrap, finishing and inspection, logistics, and the cost of delays. Here’s the simple mental math:

• Effective unit cost ≈ (Tooling or NRE ÷ quantity) + quoted unit price + finishing + shipping + expected scrap and rework.

• If a faster process pulls in your launch, the avoided delay cost can justify a higher quote price.

When quantities climb from 10 to 1000, unit price curves flatten differently by process. Support-free batch nesting makes SLS scale better than SLA at mid volumes. CNC setup time amortizes slowly but wins when tolerances bite. Rapid aluminum tooling adds a one-time cost but slashes per-unit cost above a few hundred pieces.

Benchmarks to calibrate expectations

The figures below reflect typical assumptions for a mid-complexity plastic part around 100 × 100 × 50 mm in ABS or PA12 equivalents. Real results depend on geometry, tolerances, surface spec, and supplier region.

• Protolabs noted in 2025 that SLA can turn parts in as fast as one business day and offers the smoothest surfaces among common polymers; however, its support removal and post‑processing slow scaling compared to SLS at higher counts, as discussed in their 2024 comparison of the two processes. 

• Xometry’s 2026 tolerance guide lists common SLS accuracy baselines around ±0.3% with a minimum of ±0.3 mm, helping you judge fit needs. 

• For vacuum casting, multiple primers place silicone mold life at about 20–25 shots per mold depending on geometry and resin, and typical lead times in the one-to-three‑week range after the master is prepared. 

• For rapid aluminum tooling, industry guides in 2025 commonly cite single‑cavity costs from roughly $1k–$10k with T1 build times of two to four weeks depending on complexity. 

  
  

Assumptions are generalized; calibrate with RFQs and drawings. Think of this table as the “map,” and your quote set as the “terrain.” When they disagree, the terrain wins.

DFM quick checks and tolerance sensitivity

• SLS and SLA

• SLA yields the smoothest skins and fine details but needs support removal and post‑processing, which add labor and can introduce blemishes. Protolabs outlined these trade‑offs in 2024–2025. SLS avoids supports, packs parts tightly, and delivers durable PA12 parts, but plan for a slightly grainy surface unless bead‑blasted and sealed.

• Vacuum casting

• Expect roughly 20–25 parts per silicone mold for mid‑complexity housings. Mold count, resin hardness, and demold angles move yield and cost. As summarized in 2025 primers, planning two or more molds for 100–300 parts is common practice.

• Rapid injection molding

• Draft matters. As common 2024–2025 design tips emphasize, 1–2 degrees on smooth surfaces—and more for textures—reduces ejection wear in aluminum tools. Uniform walls in the 1.5–2.5 mm range help with flow and cosmetics. Tooling NRE grows quickly with slides, lifters, or aggressive tolerances.

• CNC machining

• If your drawing implies ISO 2768 general tolerances, align expectations early. When several features need ±0.05–0.10 mm, CNC or molded parts with robust process control are the more reliable path.

Tolerance sensitivity example

• Baseline: A small enclosure needs ±0.2 mm on most features and ±0.1 mm on two mating faces for 100 units.

• Option 1 SLS path: SLS can typically hold ±0.3% (min ±0.3 mm). You’ll likely shim or post‑machine the two faces, which adds unit cost and risk.

• Option 2 CNC path: CNC meets ±0.1 mm consistently but has higher setup and per‑unit costs at 100 pieces.

• Option 3 rapid tooling path: A simple aluminum mold with draft can meet ±0.1 mm on key faces with process controls, and total per‑unit cost may undercut CNC once tooling is amortized across 100–300 units.

Result: If the ±0.1 mm requirement applies to multiple features, move sooner to CNC or molding. If it’s isolated and post‑machinable, SLS plus a machining op can still be the lowest landed cost.

Example paths that minimize total cost

Path A 10 to 50 to 300 units

• Start with SLA for looks-like approval and surface tuning; switch to SLS for functional pilots that need toughness. For the 300-piece demo run, vacuum casting delivers color and texture with predictable cosmetics and reasonable per‑unit cost, given two to four silicone molds.

Path B 20 to 200 to 1000 units

• Use CNC block‑ups for assembly fit and tolerance verification. Move to SLS for pilot assemblies to shake out geometry at scale. When the design freezes, invest in a rapid aluminum tool to hit 1000 validation units with production resins and stable repeatability.

Path C metals 10 to 150 units

• Machine aluminum prototypes quickly, then a small CNC batch with finishing such as anodizing. Only consider metal additive when geometry or internal channels mandate it; otherwise CNC will be more economical and consistent.

Supplier selection and QA checklist for bridge runs

• Share fully dimensioned drawings with tolerance classes and critical-to-quality notes. Include draft expectations for molding and any texture or color targets.

• Ask for DFM feedback on undercuts, draft, thin walls, and gating or ejection features if molding is in play.

• Request sample inspection data on the first articles and define an acceptance plan for the bridge lot.

• For mixed paths, confirm who owns the master model and whether it can be reused across SLA, SLS, and casting.

• When moving to tooling, align on resin, shrinkage assumptions, and cosmetic grade early to avoid rework.

If you’re evaluating CNC for tolerance-driven parts, it can help to review a vendor’s capabilities and materials up front. For example, see the contextual overview on CNC machining at Kaierwo and their CNC plastic machining notes when plastics are involved. If your roadmap points to urethane casting or a tooling handoff, these primers outline process steps and typical timelines: vacuum casting service and rapid tooling. These internal references are provided as neutral context to help you scope conversations with any qualified supplier.

A practical note on execution: If you prefer a single provider that supports SLS, vacuum casting, CNC, and rapid tooling under one roof or via vetted partners, consolidating steps can compress quoting and logistics. A shop like Kaierwo, for example, publishes public guides on SLS‑to‑vacuum casting handoffs and rapid tooling timelines; linking these steps inside one workflow can reduce changeover friction while keeping a neutral, data‑first vendor selection mindset.

Author’s note This guide was written from the perspective of a manufacturing engineer focused on short‑run hardware builds. It blends public vendor benchmarks with common DFM rules of thumb. Always let your quote set and first-article inspection data drive the final call.