HomeProductsModel Catalog

Model Catalog

Nobody wants an empty GPU; they want something that runs. A template is one image plus sane defaults — disk, ports, a recommended GPU. The "pick one, start" flow on the home page and the console's rent dialog are fed by this same list.

4 ready 4 preparing Custom image always works

Console links point to the console domain; when you open the site locally they are rewritten to localhost:3000.

Where things stand

Four ready, four preparing.

Our status rule is one sentence: a template gets called ready when it runs today from a public, official image, and nothing else gets called ready. We would rather keep the list short than put a "ready" label on something that is not.

Ready — 4 templatesJupyter + PyTorch · vLLM · Ollama · Axolotl
Preparing — 4 templatesComfyUI · Whisper · Blender · Agent runtime
Custom image — always worksPaste the registry address, the pod comes up
Signature verification (cosign) and image pre-pullPhase 2

A green dot here means "written as part of Phase 0 and running in local development" — it is not a statement about production health. Component by component: Status.

Starting templates

Ready-made pods: pick one, start

There is no price column: the price follows the GPU you pick, not the template. The recommended GPU below is a sensible starting card for that kind of work; current hourly rates are on the pricing page.

TemplateWhat it doesCategoryRecommended GPUMin VRAMStatusStart
Jupyter + PyTorch Experiment and prototype; from a notebook to a training script dev RTX 4090 24 GB ready Start →
vLLM · OpenAI-compatible server Publish a model as an API; only base_url changes on the client inference A100 80GB 48 GB ready Start →
Ollama · your own chat Pull and run open models with a single command inference RTX 4090 24 GB ready Start →
Axolotl · fine-tuning LoRA/QLoRA and full fine-tune training A100 80GB 48 GB ready Start →
ComfyUI · image generation SD/SDXL/Flux workflows, node-based interface image RTX 4090 24 GB preparing Start →
Whisper · speech transcription Transcription, subtitles, batch audio processing audio RTX 3090 16 GB preparing Start →
Blender · render 3D render, animation, architectural visualization render RTX 4090 24 GB preparing Start →
Agent runtime Tool-calling LLM agents, long-running jobs agent RTX 4090 24 GB preparing Start →
Custom image Your own container — paste the registry address dev ready Start →

preparing means: the template package is still being put together and the image is not pinned yet. You can run that work today anyway — "Start" hands you the custom-image flow, and the only difference is that you supply the image. We deliberately do not print image tags on this page: tags move, pages go stale, and the current image sits in the console's rent dialog.

Categories

What can you run?

A category is not just a label: the defaults behind it decide which ports the pod opens, how much disk it asks for and which card is a sensible starting point. Where a template is still preparing, the work itself is possible today anyway — with your own image.

dev — a place to work

Jupyter + PyTorch for the notebook-to-script path, and the custom image for everything the catalog does not cover. Both run today.

readyjupyter · custom

inference — serving a model

vLLM turns an open model into an OpenAI-compatible API; Ollama is the quickest path to your own chat. Detail: Inference.

readyvllm · ollama

training — teaching the model

Axolotl covers LoRA, QLoRA and full fine-tunes; checkpoints belong on a volume that outlives the pod. Detail: Fine-tuning.

readyaxolotl

image — generation workflows

ComfyUI's node graph for SD/SDXL/Flux. The template is preparing; today the same workflow runs from your own image. Detail: Image & video.

preparingcomfyui

audio — speech to text

Whisper for transcription, subtitles and batch audio jobs. The template is preparing; today it runs from your own image, and a 16 GB card is enough.

preparingwhisper

render — frames, not tokens

Blender for 3D render, animation and architectural visualization. The template is preparing; today it runs from your own image.

preparingblender

agent — tool-calling runs

An agent runtime for LLMs that call tools and jobs that stay up for hours. The template is preparing; today it runs from your own image.

preparingagent

Not in the catalog

Multi-node cluster training belongs to Clusters — after Phase 2, soon. A scale-to-zero production endpoint belongs to Serverless Inference — Phase 2, soon. Neither is a template, and neither exists today.

soonnot a template
Whichever template you pick: per-second billing, a hard budget cap, and you don't pay for seconds that don't work.
The scorecard picks the machine: the template says what runs, the scorecard says where it runs — price × reliability × network.
Region pinning still holds: a template does not move your data — the pod stays in the country you pinned it to.

What a template is

An image, a few defaults, one less decision.

A template does not put a layer between you and the GPU. It answers the three questions you would otherwise answer by hand before the first log line.

The image and its defaults travel together

Every template carries a volume size, the ports it needs and the card it is comfortable on. You can override all of them in the rent dialog; the point is that the defaults are not a guess you have to make at midnight.

"Start" is a link, not a wizard

Each Start above opens the console's rent dialog with that template preselected. If a link ever carries an id we no longer recognise, it is ignored quietly and the screen still opens — a stale bookmark should not break a page.

A recommended GPU is a floor, not a rule

The minimum VRAM figures are what the template comfortably fits into. The real choice follows your model size and context length; the marketplace will happily rent you something larger, and the pricing page shows what that costs. On a stopped pod, storage still bills — that line is visible on the billing screen.

Signing and pre-pull are the catalog's real job

Ready-made templates are a convenience. Verifying that a catalog image is the one we published, and having it already sitting on the host, is the part that touches security and money. Both are Phase 2 work, and neither is running today.

Pre-pull is a price feature, not a speed feature: pulling ten gigabytes while the pod boots disappears into the noise of an hourly bill, but on a per-second bill it comes straight out of your pocket. Pre-pull and the model cache both need persistent space on the host side, which depends on our own host network — Phase 1 and Phase 2 in the technical plan. The same mechanism is what would make the cold-start target of Serverless Inference possible.

From code

The catalog is an endpoint, not a landing page.

The same list feeds this page, the console's rent dialog and your own scripts. If you would rather stay in the terminal, none of this needs a browser.

Pin your own version: write the image tag out in full and stay away from "latest".
Keep the image thin: until pre-pull lands, every megabyte is start-up time, and start-up time is on the bill.
Checkpoint onto a volume: a pod is disposable, your work is not. See Persistent Storage.
CLI
# The catalog comes from the API — output trimmed, illustrative $ curl -s localhost:8080/v1/templates {"templates":[{"id":"vllm","category":"inference", "status":"ready","recommended_gpu":"A100 80GB", "min_vram_gb":48,"volume_gb":100,"ports":[8000]}, …]} # A preparing template today: same pod, your own image $ kaldera pod create --offer ofr-8817 \ --image ghcr.io/your-team/comfy:1.4.2 \ --volume-gb 200 ✓ pod ready → ssh root@fra1.kaldera.ai -p 40122

Contribute

Which template would you want next?

Which four get packaged next is decided by the work people actually run. Tell us what you run, and the list gets better rather than longer.

The mailbox is a placeholder and opens soon.