VHH, scFv or IgG: which format should I choose for my application?

Written by the Immunochem scientific team and reviewed by Philippe Herman, CEO and Chief Scientific Officer. Published 13 October 2026, updated 13 October 2026.

Which format should I choose for my application?

Start from the job, not from the molecule. Four questions settle most cases.

How long does it need to circulate? A VHH or scFv is cleared by the kidney within hours. An IgG persists for weeks. If dosing frequency matters, that single fact outweighs everything else.

Does it need to kill or recruit? Effector functions such as ADCC and complement activation come from the Fc region. No Fc, no effector function. Fragments can be fused to an Fc to get it back, which is a common and well understood move, but then you are no longer choosing between the formats, you are building one from the other.

Where does it have to reach? Dense tissue, a tumour core, or an epitope at the bottom of a cleft all favour the smallest molecule that will do the job.

What are you building? If the end product is a bispecific, a CAR construct, a multivalent binder or a diagnostic reagent, you are choosing a building block rather than a drug, and the single-domain format is simpler to assemble.

The three formats side by side

The comparison between the two fragment formats is covered in detail in a review by Asaadi and colleagues in Biomarker Research, which reports higher solubility, higher stability, smaller size and lower production cost for VHHs (DOI, reference located through PubMed). Their conclusion is not that one format wins, but that the two suit different applications, which matches what we see in practice.

What size buys you, and what it costs

A VHH is roughly a tenth the mass of an IgG, and that single difference drives most of the others.

Small means it reaches places a full antibody does not. It crosses dense tissue, it penetrates tumours more evenly, and it can engage epitopes that sit in a groove or a cleft where there is no room for a large paratope. Enzyme active sites and receptor binding pockets fall into this group, and it is a real functional advantage rather than a convenience.

Small also means it leaves quickly. Below roughly 60 kDa a protein is filtered by the kidney, so a VHH or an scFv is cleared in hours. For an inhaled product, a topical product, or an imaging agent, that is exactly what you want, since you do not want the tracer to stay. For a systemic therapy dosed every few weeks, it is the problem to solve, and it is solved by fusing to an Fc, to albumin, or to an anti-albumin binder.

The second consequence of size is that a single domain carries no chain pairing. An scFv holds its VH and VL together with an artificial linker, and that linker is the origin of most of the format's problems: domain swapping, linker-dependent aggregation, and batch variability. A VHH has nothing to come apart.

Half-life and effector function: the case for IgG

If your product needs to circulate for weeks and recruit immune effectors, an IgG is the right molecule and no amount of engineering on a fragment gets you there as directly. The Fc gives you both properties at once, through FcRn recycling for persistence and Fcγ receptor engagement for effector function, and the regulatory and manufacturing path for IgGs is the best understood path in the industry. That is worth something real when you are planning a programme rather than a proof of concept.

A fragment-plus-Fc fusion recovers much of this, and plenty of approved products are built that way. But it is a bigger molecule with more to characterise, and if you were always going to need an Fc, starting from an IgG is the shorter road.

When each format is the right answer

Choose a VHH for intracellular and intrabody work, imaging and diagnostics, inhaled and topical delivery, recessed or cryptic epitopes, CAR and bispecific building blocks, research reagents that must survive handling, and any case where you need high stability in a small package.

Choose an scFv when you are working inside a platform that already runs on scFvs, when you need a bivalent fragment without an Fc, or when an existing scFv-based construct is the thing you are extending. One caveat if your application is CAR work: a 2026 comparison from Autolus and University College London built ten matched VHH and scFv pairs against CD123 and found the VHH versions close to 8 °C more thermostable, less aggregation-prone, and, as CAR domains, better at reducing tumour burden and extending survival in a mouse model of acute myeloid leukaemia (Kinna et al., Molecular Therapy, 2026, DOI, reference located through PubMed).

Choose an IgG when you need weeks of circulation, when you need effector function, when you are following a well-trodden regulatory path, or when the target biology requires the avidity of two binding sites held at a fixed distance.

What the format choice does to the design work

This is the part that rarely appears in format comparisons, and it matters if the molecule is going to be designed rather than discovered.

A single domain is a smaller design problem. One chain, no pairing interface, fewer positions whose only job is to hold the molecule together. That means more of the sequence is available to vary, which matters whenever you are designing under a constraint, for instance when you need to stay a measured distance from a patented sequence. It also means a shorter design cycle: our working timelines are six weeks to candidate sequences and twelve weeks to binders validated in the laboratory, and they are shorter for a single domain than for a paired format.

So the honest summary is that format choice is driven by your application, and the design cost is a secondary consideration. But if two formats would both work for your application, the single domain is the cheaper one to design and the faster one to test.

Frequently asked questions

Is an scFv the same thing as a VHH?

No. An scFv is two domains, a VH and a VL, joined by an artificial linker. A VHH is one naturally single domain, derived from heavy-chain-only antibodies. The scFv is an engineered construct; the VHH is a natural architecture.

Can a VHH replace an IgG in a therapeutic?

Sometimes, and not as a direct swap. You will need to address half-life, usually by fusing to an Fc or to albumin, and you will have no effector function unless you add it. For local delivery, imaging, or inhibition where effector function is not wanted, a VHH can be the better therapeutic.

Which format has the highest affinity?

None of them, as a property of the format. Affinity is a property of the individual binder. Single domains reach picomolar affinities, and so do IgGs. The formats differ in stability, size and half-life, not in the affinity ceiling.

Do I have to decide before starting?

For the design work, mostly yes, because the format sets the scaffold. But a VHH can be reformatted afterwards into a bivalent construct, an Fc fusion or a CAR domain without redesigning the binder, which is one of the reasons it is a convenient place to start when the end format is not yet settled.

Do you work in all three formats?

We design single-domain antibodies and conventional antibodies. We do not run animal immunisation for any of them: the sequences come from computational design against the target structure, and we validate binding in the laboratory before delivery.