I have not used the H3C or this feature, so this guide is built from Bambu’s wiki, its Bambu Studio release notes, its forum and store pages, and the independent reviews published since launch. I mark which is which, and I flag where something is my own reasoning.
Among the H3C’s launch features, the one I think has the most long-term potential is also the least glamorous: it can print a single model with different nozzle sizes. A fine 0.2 mm nozzle for the text on a dial, a 0.4 mm nozzle for the case, and a larger one for bulk, all in one job and each at its own layer height. Printing with more than one nozzle size is not new, but doing it this way is, and it goes straight at the oldest trade-off in the hobby: speed against detail. Here is how it works, how to set it up, and where it falls short.
The short version
| Question | Answer |
|---|---|
| What is new? | Bambu Studio can assign different nozzle diameters to different regions of one model, and each nozzle prints at its own layer height and line width |
| Which printer? | The H3C only for now. Bambu Studio 2.9.1 (3 November 2026) added it |
| Other printers? | Bambu says H2D, H2C and X2D are coming: H2D in late November 2026 (without Filament Track Switch), H2C and X2D in February 2027, with Filament Track Switch setups a month or two later |
| What do you need? | Extra hotends, since only 0.4 mm ones are included, and a model where the detail and the bulk are in different places |
| Is it automatic? | No. You assign nozzles to regions yourself with a filament setting, a modifier or the colour painting tool |
| Is it always faster? | No. Bambu itself says the answer depends on the model |
The idea: layer height belongs to the nozzle
A nozzle can only lay down sensible layers within a range of its diameter. Bambu’s layer height guidance puts the reasonable range at about 20% to 70% of the nozzle diameter, so for a 0.4 mm nozzle that is roughly 0.08 to 0.28 mm. Fine layers capture detail but cost time. Bambu’s wiki shows the cost using the same model printed three ways.
| Layer height (0.4 mm nozzle) | Print time | Compared with 0.2 mm (my arithmetic) |
|---|---|---|
| 0.08 mm | 1 h 19 min | About 2.2 times as long |
| 0.20 mm | 36 min 14 s | Baseline |
| 0.28 mm | 28 min 27 s | About a fifth less time |
Two things stand out. Going from 0.2 to 0.08 mm multiplies the layer count by 2.5 but costs only 2.2 times the time, and going from 0.2 to 0.28 mm saves a good deal less than the layer count alone would suggest. Time is not a simple function of layer height, so the real savings from any mixed-nozzle trick will be smaller than a naive estimate. The point still stands: if one nozzle has to serve the whole model, either the bulk is printed unnecessarily finely or the detail is printed unnecessarily coarsely.
What was possible before, and what is different
- Assigning nozzles by line type. Slicers have long let you give walls, infill and supports different nozzles, and Bambu Studio did too, through the filament settings under Process, Strength and Advanced. That is useful, but accuracy needs usually follow surface design rather than line types. Text and the smooth back of a model are both “outer wall”.
- Splitting the model by hand. You could cut a model into separate parts, give each its own nozzle and layer height, and keep them aligned. Bambu’s wiki says this becomes tedious with interlocking soft and hard materials, curved boundaries or surface colour painting, and every edit means splitting again.
- One shared layer height. Most slicers handle mixed diameters by forcing everything to share a layer height, which means either slow, thin layers for the large nozzle or coarse layers for the fine one.
The new capability follows model regions rather than line types, and lets each nozzle use its own layer height even within the same height range. Bambu says this needed a deep refactor of the slicing architecture, because traditional slicers organise a model around global Z heights, and neighbouring regions growing at different rates break that assumption. The slicer must also decide the order of tool changes and travel moves so the nozzle does not scrape formed structures. Bambu says all multi-nozzle models benefit from the refactor, but the H3C is supported first.
How to set it up, step by step
This follows Bambu’s wiki guide. I have kept to what it says and added only a few notes of my own.
- Fit the nozzles you need. The H3C comes with 0.4 mm hotends only, and the other sizes (0.2, 0.6 and 0.8 mm) are sold separately. Each filament must sit in the tool whose nozzle suits it. The printer detects the hotend type automatically, and if you swap one, use Refresh or Refresh All on the screen or in Studio’s Device page.
- Sync the printer. Connect in Bambu Studio, go to the Prepare page and click Sync, so Studio picks up the nozzle types and AMS units.
- Switch to multi-diameter mode. Studio lets you switch between multi-diameter and single-diameter modes. In multi-diameter mode you can assign each project filament to one of the available nozzle types.
- Sync the AMS filaments. Use Project Filament Sync. There are two matching methods, Accuracy Priority (matches nozzle diameter first, then colour) and Color Priority (the reverse), plus an Overwrite mode. Bambu warns that matching can change a filament’s nozzle diameter, which changes the model’s accuracy, so check the result.
- Set the regions. There are three ways: set the nozzle for a filament directly in the project filament list, add a modifier and give it a different filament and nozzle combination, or paint the model with the colour painting tool after assigning nozzles to filaments.
- Choose the process parameters. Pick a quality profile by base layer height. Studio works out the other nozzles’ layer heights and line widths from it (see the next section).
- Slice, then inspect. Layer height and line width should vary by diameter in the preview. Zoom into the region boundaries and the support gaps.
- Send the print. The filament mapping is strictly tied to nozzle diameter. A filament sliced for a 0.2 mm nozzle can only print through a 0.2 mm nozzle.
The layer height rule
You cannot set layer heights for each nozzle freely. Bambu’s wiki gives the formula: nozzle layer height equals the preset layer height multiplied by the nozzle diameter and divided by 0.4. The base is the layer height of the 0.4 mm nozzle, and Studio fills in the rest. Here are some examples, the first of which comes straight from the wiki.
| Base layer height (0.4 mm) | 0.2 mm nozzle | 0.4 mm nozzle | 0.6 mm nozzle | 0.8 mm nozzle |
|---|---|---|---|---|
| 0.12 mm (from the wiki) | 0.06 mm | 0.12 mm | 0.18 mm | 0.24 mm |
| 0.20 mm | 0.10 mm | 0.20 mm | 0.30 mm | 0.40 mm |
| 0.24 mm | 0.12 mm | 0.24 mm | 0.36 mm | 0.48 mm |
A pattern emerges when you work it through, and this is my own observation, not Bambu’s. Every nozzle prints at the same fraction of its diameter, so each runs at the same relative squish and the same quality “feel”. Because the four sizes are in the ratio 1:2:3:4, the layer heights are too, so on the 0.20 mm base a 0.4 mm layer from the 0.8 mm nozzle equals exactly four 0.1 mm layers from the 0.2 mm nozzle. That is presumably why the rule is a strict ratio: it lets thick and thin layers meet at common heights.
One quirk to know about. If your initial layer height differs from the regular layer height, the larger nozzle’s actual first layer may come out lower than expected. Bambu’s example is a 0.12 mm preset with a 0.2 mm first layer, where the 0.4 mm nozzle actually prints 0.18 mm. Its recommendation is to set the initial layer height equal to the regular layer height.
Why it is a good thing
- It attacks the real trade-off. The fine nozzle only works where detail is needed, and a larger nozzle does the bulk. TechRadar describes the use: a fine 0.2 or 0.4 mm nozzle for the exterior and a 0.6 or 0.8 mm one for interior structure where speed matters more.
- It follows the design, not the line type. Detail regions get detail settings. That is what a designer wants, and it is what splitting a model by hand tried to achieve.
- It combines with materials. Bambu’s showcase is the Maker Watch, a dial, case and strap in one print: a 0.2 mm nozzle for the dial’s tiny numbers, a 0.4 mm nozzle for the case, and a 0.4 mm nozzle on the 2.85 mm tool for a TPU 75A strap. I covered that filament in my TPU 75A post. Bambu also shows a hinge lamp with a 75A TPU damper printed between rigid shells, with a 0.2 mm nozzle for the ratchet.
- It can be big on the right model. A forum user who split a ship model by hand on an H2D reported that printing the fine sections at 0.2 mm and the rest at 0.4 mm gave a total of 25 h 28 min against an estimated 52 h 22 min for the 0.2 mm approach, about half. That is one user’s hand-split, not a controlled test, and a true mixed print would add switching overhead, but it shows the potential.
- It spreads. Bambu says it has built this for the H2D, H2C and X2D too, so the feature is not a one-printer gimmick.
When it is worth using
Bambu’s wiki gives two tests. First, how large a share of the model is detail? The more concentrated the detail and the bigger the main body, the more worthwhile it is. Second, over how many layers do the nozzles share the work? Where both are active in the same height range, switching adds time and extra walls and material at the boundaries.
| Good fit | Poor fit |
|---|---|
| An enclosure with small markings or text on one face | A model that is fine detail almost everywhere |
| A functional part with slanted surfaces or engraved details | A small model, where the fine nozzle still sets most of the time |
| A large body with one small detailed region | Detail scattered through the full height, so the nozzles alternate on most layers |
| Soft and rigid parts with different nozzle needs, such as the watch | Anything where you would not notice the difference in layer height |
The cost of interleaving is easy to underestimate. CNC Kitchen measured tool changes on the H3C at about 6 to 7 seconds each, with about 50 mg of filament going to the prime tower per change. By my arithmetic, a thousand changes would cost roughly 1.7 to 1.9 hours and about 50 g of filament. A model where both nozzles work across many layers can rack that up quickly.
The honest limits
| Limit | Detail | Source |
|---|---|---|
| Regions are manual | Studio does not detect detail automatically. You assign nozzles by filament, modifier or painting | Bambu wiki |
| Layer heights are ratio-locked | You cannot choose them freely per nozzle | Bambu wiki |
| Some parameters are automatic and uneditable | Studio calculates them from the model and the nozzle diameter, and users cannot customise them yet | Bambu wiki |
| Speed and acceleration | Early testing found they could not always be tuned independently per nozzle, and overhang settings had similar limits | Makers101 |
| Support gaps | Automatic handling of support Z gaps is not fully optimised. Inspect after slicing | Bambu wiki |
| No per-layer filament changes | In multi-diameter mode, changes added by layer in the preview are lost. Use a height range modifier instead | Bambu wiki |
| Boundaries | Gaps or surface differences can show where regions meet, especially on slanted or gentle interfaces. Test print critical models | Bambu wiki |
| Prime tower still needed | A forum tester saw it with default settings, along with an odd 0.4 mm bridging layer over the infill in one test | Bambu forum |
| Duplicate spools | A single-colour model that needs fine detail needs two spools of the same colour, one in the 0.2 mm slot and one in the 0.4 mm slot | Bambu wiki |
| Speed gain is not guaranteed | In one 0.2 and 0.4 mm test, a small mixed-nozzle model took about as long as the all-0.2 mm version once the prime tower was included | Makers101 |
| Workflow | Hackster’s reviewer dislikes the automatic approach and the inability to set layer thicknesses for larger nozzles, and hopes for something better | Hackster |
A forum commenter put it bluntly: it is not plug and play, and needs modelling skill as well as slicing skill. That seems fair for a first release. SimplyPrint quotes one reviewer’s verdict, that it “still feels like a technology in development, the potential is obvious”, which is how I would describe it too.
What it costs
| Item (US store) | Price |
|---|---|
| Multi-Size Hotends Kit, H3C (seven hotends, in 0.2 mm, 0.4 mm or 0.6 mm sets, or a mix of 0.2 x 4, 0.6 x 2 and 0.8 x 1) | $244.93, shown as $205 when added to an H3C order |
| Single Bambu Hotend, H3C | $29.99 (typical $34.99) |
| Bambu High Flow Hotend, H3C | $59.99 |
| Bambu Hotend for 2.85 mm filament, H3C | $29.99 (typical $34.99) |
Bambu says the hotends swap with three screws in under a minute. The Maker Watch needs three 0.2 mm hardened-steel hotends, sold separately. One caution from Bambu’s own H2-series hotend page, not H3C-specific: abrasive carbon and glass fibre filaments are prone to clog a 0.2 mm nozzle, and it recommends a hardened 0.6 mm nozzle for them. My abrasive filament guide and the nozzle types guide cover the background.
A sensible way to start
- Pick a model that clearly suits it, such as a large part with one small detailed area.
- Put the detail filament in the 0.2 mm slot, and supports or infill in a 0.6 or 0.8 mm slot, as Bambu suggests.
- Set the initial layer height equal to the regular layer height.
- Use a height range modifier if you want a filament change at a specific height.
- Compare the slicer’s time estimate for the mixed print against a single-nozzle slice before you commit.
- Zoom into the preview at every boundary and do a small test print before a long one.
When other printers get it
| Printer | Without Filament Track Switch | With Filament Track Switch |
|---|---|---|
| H2D | Late November 2026 | February 2027 |
| H2C | February 2027 | March 2027 |
| X2D | February 2027 | March 2027 |
Bambu’s forum post says each of these printers has a different AMS setup from the H3C, so the slicing logic and parameters need adapting, and Filament Track Switch setups need more work. Plenty of owners were unhappy about the wait, especially H2C owners, who are at the back of the queue. That is a fair reaction, though I cannot tell you whether the wait is technical or commercial. The timeline is Bambu’s estimate, not a commitment. The A1 and A2L, which I own, are not on the list.
What I would want tested
- A like-for-like benchmark: the same model printed all at 0.2 mm, all at 0.4 mm and mixed, with time, filament use and surface quality recorded.
- How much the region boundaries show on real prints, especially on slopes.
- How the prime tower and tool-change overhead scale as the nozzles alternate more often.
- Whether speed and acceleration limits per nozzle become independently tunable in later releases.
- How well it works with fibre-filled filaments and the 0.2 mm nozzle.
- Whether MakerWorld profiles will show nozzle requirements clearly, as a forum user asked.
My take
I think this is a real step forward in concept, and a modest one in execution so far. Concept first: letting the design decide where fine detail lives, and giving each nozzle the layer height it is suited to, is the right approach, and it is the kind of problem that has been worked around rather than solved for years. Execution second: it is manual, it is rigid in its rules, it still needs a prime tower and it is only on one printer, so the benefit will depend heavily on the model. For a part like the Maker Watch, or an enclosure with a small engraved logo, it could be a big win. For anything else, check the slicer estimate first. The refactoring behind it is probably the most important part, because it is what will let the same idea reach more printers and more settings. I will update this guide as testing and the H2D release arrive.
Sources and further reading
- Bambu Lab wiki: H3C mixed nozzle size printing guide
- Bambu Studio release notes (2.9.1, H3C support and mixed nozzle slicing)
- Bambu Lab forum: mixed nozzle size printing is coming to H2D, H2C and X2D
- Bambu Lab: H3C overview
- Bambu Lab US store: H3C, hotend kits and the Maker Watch note
- Bambu Lab US store: Multi-Size Hotends Kit, H3C
- Bambu Lab wiki: layer height in Bambu Studio
- Hackster: The Maker’s Toolbox, Bambu Lab H3C review
- TechRadar: Bambu Lab H3C review
- CNC Kitchen: Bambu Lab H3C review (tool change times and prime waste)
- Makers101: Bambu Lab H3C price, specs and early tests
- SimplyPrint: Bambu Lab H3C reviews summary
- ToolGuyd: Bambu teases multi nozzle compatibility on H3C
- Bambu Lab forum: user’s hand-split H2D time comparison
