Where each model runs

Which bitHuman model runs where: the four product models, what GPU-only means, and which one to pick.

This is the page to read before you pick a model. It answers one question — which model runs where — and it does not soften the answer anywhere.

The four models

Those four are the only product names. If you have met the words elevate, embody, essence-2-light, essence-2-quality, lebundle or libelevate, see legacy names you will still see — several of them are still literals you have to type or read, and this page shows you which.

The matrix

There are two answers, and they are two different facts

In scope — this model belongs on this platform. If the artifact is missing here, that is a gap, and it is our bug.

Not applicable — this model is deliberately off this platform. A missing artifact here is correct. It is not “coming soon”, it is not a roadmap item, and there is no date. expression-1 is GPU-only by design: its absence from Apple, the browser and Android is the intended shape of the product, so architect around a GPU for it rather than waiting. If you need photoreal quality on a Mac, a phone or in a tab, the model you want is essence-2.

essence-1’s Not applicable cells are the other kind: they are the cloud’s own serving tiers, and we serve essence-1 from the cloud’s Apple tier. That is about where we run it. Nothing about where you run it changes — your Mac, iPhone, browser and Android stay in scope.

essence-1 is the model you are most likely to be handed

About two thirds of the agents on the platform are essence-1, and the CLI’s default showcase catalogue is built entirely from it — every avatar a bare bithuman list returns is an essence-1 identity. A second catalogue of twenty essence-2 and expression-2 identities is open to anyone with no credential; the CLI reads it with bithuman list --manifest https://api.bithuman.ai/v1/models/showcase.

One thing to code against: a stored agent can carry no model value at all. Roughly a fifth of them do, so anything of yours that switches on agents.model must handle a null. It is not a rounding error.

Identify what you are holding

bithuman info reads a model file locally and tells you its family. It needs no credential and no network, so it is the cheapest way to find out which row of the matrix applies to a file somebody sent you:

bithuman info e2.lebundle.imx

  Engine:         essence2-light
  Family:         essence-2 (Essence 2)
  Members (26):
    …

Read the Family: line — that is the product. Two things in that output are legacy names you will meet and cannot avoid: the file extension .lebundle.imx, and the Engine: essence2-light line. Neither is a product name; legacy names maps them all. The full output and every exit code are on the CLI reference.

What a credential changes

bithuman info is free. bithuman render is not — it needs bithuman login, or BITHUMAN_API_SECRET in the environment. Get a key at Developer → API keys.

Without one, render refuses before it opens the file, so the refusal is the same for every model family. bithuman doctor reports the same thing as a checklist and exits non-zero on a host with no credential — that is the designed result, not a broken install.

The lanes, one at a time

Measured frame rates for every platform are on the performance page.

Legacy names you will still see

There are two current product names — expression-2 and essence-2 — plus the first generation, essence-1 and expression-1. elevate, embody, essence-2-light, essence-2-quality, lebundle and libelevate are deprecated as words.

Deprecating a word does not rename a wire format. Several of these are frozen forever in file names, API fields and manifest values, and you will have to read or type them. Hiding a name you have to type would be worse than showing a retired one, so here they are:

The rule: write the product name; accept the legacy spelling on input; expect to read it in file names and engine strings forever.

One more naming point, because it causes real architecture mistakes: the Apple lane is called Apple, not “ANE”. It is the whole Apple Silicon target, not one accelerator inside it — we run the work on whichever unit measures faster.

Which model should I use?

You need it on a Mac, an iPhone, in a browser, or on Android. essence-1, essence-2 or expression-2. expression-1 is GPU-only and that is permanent — plan for the cloud, not for a future release.

You want photoreal, everywhere. essence-2.

You need maximum concurrency on cheap hardware, or a 24/7 unattended display. essence-1. Low memory, no idle timeout, custom gestures, runs on 1-2 CPU cores.

You want to supply the face at session time, or the character is stylized or non-human. expression-2.

You are maintaining an existing v1 agent. expression-1 on GPU, or essence-1 on any lane you run yourself. Both remain supported.

Next steps

• Essence 2 & Expression 2 — the second-generation family overview
• Essence 2 · Expression 2
• Essence vs Expression — the first-generation pair in detail
• CLI reference — every command, flag and exit code
• Android SDK · Swift SDK · Python SDK · CLI
• Avatars and the .imx format — how a model file is packaged
• Pricing & credits — what each model costs to run