How to run LiteLLM Proxy in front of a self-hosted model on a cloud GPU

Rent a GPU on Deploy, then open a terminal in JupyterLab or over SSH. Install LiteLLM Proxy the way the official repository prints, serve an open-weight model on that machine, and point a model_list entry at that server. Copy the proxy config and the .env that holds the master key before you destroy the instance, because destroying the instance deletes its disk and stops both the proxy and the model server.

An agent can rent the GPU, serve the model, and share that one gateway with other agents. The agent pages are GPU compute for AI agents and the agent integration guide. Those pages are how an agent discovers a GPU and runs work on it. This page is the LiteLLM Proxy setting. A TypeScript SDK whose compatible-providers page calls createOpenAICompatible is How to use the Vercel AI SDK with a self-hosted model on a cloud GPU. An agent whose local sample sets base_url to http://localhost:11434/v1 is How to run Semantic Kernel agents with a self-hosted model on a cloud GPU. An answering engine whose Docker command publishes port 3000, and whose Ollama base URL placeholder is http://localhost:11434, is How to run Perplexica AI search with a self-hosted model on a cloud GPU. An assistant whose setup page opens http://localhost:42110, and whose Ollama page sets OPENAI_BASE_URL to http://localhost:11434/v1/, is How to run Khoj AI assistant with a self-hosted model on a cloud GPU. An editor whose Ollama config sets baseURL to http://localhost:11434/v1 is How to use Kilo Code with a self-hosted model on a cloud GPU. An editor whose local-model page sets the Ollama api_url to http://localhost:11434 is How to use Zed’s AI agent with a self-hosted model on a cloud GPU. A coding assistant whose Docker command publishes port 8080 and passes --model and --chat-model is How to run Tabby self-hosted coding assistant on a cloud GPU. A terminal client whose OpenAI-compatible page sets OPENAI_API_BASE is How to use Aider with a self-hosted model on a cloud GPU. An editor client whose install page is the VS Code extension is How to use Cline with a self-hosted model on a cloud GPU. An editor whose config.yaml sets an apiBase for chat, autocomplete, and embeddings is How to use Continue with a self-hosted model on a cloud GPU. An editor whose Ollama provider defaults to http://localhost:11434 is How to use Roo Code with a self-hosted model on a cloud GPU. A coding agent whose providers page sets package to the OpenAI-compatible provider is How to use OpenCode with a self-hosted model on a cloud GPU. A platform whose single-container guide sets CHAT_BASE_URL to http://host.docker.internal:11434/v1 for a local OpenAI-compatible server is How to run AutoGPT with a self-hosted model on a cloud GPU. An agent whose Docker command publishes port 5080, and whose Ollama provider sets api_base to http://host.docker.internal:11434, is How to run Agent Zero with a self-hosted model on a cloud GPU. An agent whose Ollama example sets base_url to http://localhost:11434/v1 is How to run OpenManus with a self-hosted model on a cloud GPU. A chat app whose Docker page runs docker compose up -d and whose Ollama page sets baseURL to http://localhost:11434/v1/ is How to run LibreChat agents with a self-hosted model on a cloud GPU. A chat app whose README runs docker compose up -d and whose environment page sets OLLAMA_PROXY_URL to http://127.0.0.1:11434 is How to run LobeChat agents with a self-hosted model on a cloud GPU. A terminal agent whose install page runs curl -fsSL https://www.openinterpreter.com/install | sh, and whose models page sets the Ollama base URL to http://localhost:11434/v1, is How to run Open Interpreter with a self-hosted model on a cloud GPU. A workspace app whose Docker quickstart publishes port 3001, and whose Ollama page sets the URL to http://127.0.0.1:11434, is How to run AnythingLLM agents with a self-hosted model on a cloud GPU. A repository agent whose local-models section sets api_base to http://localhost:11434 is How to run SWE-agent with a self-hosted model on a cloud GPU. A workflow editor whose Docker install publishes port 5678, and whose Ollama credentials default the Base URL to http://localhost:11434, is How to run n8n AI agents with a self-hosted model on a cloud GPU. A CLI whose providers page sets OLLAMA_HOST to http://localhost:11434 is How to run the goose agent with a self-hosted model on a cloud GPU. A canvas whose self-host page opens http://localhost/install is How to run Dify with a self-hosted model on a cloud GPU. A canvas whose ChatOllama page names a base URL is How to run Flowise with a self-hosted model on a cloud GPU. A flow editor whose Ollama component defaults the base URL to http://127.0.0.1:11434 is How to run Langflow with a self-hosted model on a cloud GPU. A CLI whose connect help sets --base-url to http://localhost:8000/v1 for openai-compatible is How to run Letta agents with a self-hosted model on a cloud GPU. A browser chat whose serve command listens on port 8080 is How to run Open WebUI on a cloud GPU. A Python agent library whose CodeAgent can take TransformersModel or an OpenAI-compatible server is How to run smolagents with a self-hosted model on a cloud GPU. A crew whose LLM class takes base_url for Ollama or an OpenAI-compatible server is How to run CrewAI with a self-hosted model on a cloud GPU. A graph runtime whose overview example does not name a model is How to run LangGraph agents with a self-hosted model on a cloud GPU. An AgentChat library whose Ollama example sets model to llama3.2 is How to run AutoGen agents with a self-hosted model on a cloud GPU. A typed agent whose vLLM example sets base_url to http://localhost:8000/v1 is How to run Pydantic AI agents with a self-hosted model on a cloud GPU. A framework whose README sets Ollama and HuggingFaceEmbedding is How to build LlamaIndex agents with a self-hosted model on a cloud GPU. A browser agent whose models page sets base_url to http://localhost:11434/v1 is How to run browser-use with a self-hosted model on a cloud GPU. A pipeline whose OllamaChatGenerator sets url to http://localhost:11434 is How to run Haystack agents with a self-hosted model on a cloud GPU. A program whose language-models page sets dspy.LM to ollama_chat/llama3.2:1b is How to run DSPy with a self-hosted model on a cloud GPU. An agent whose Ollama example sets id to llama3.1:8b is How to run Agno agents with a self-hosted model on a cloud GPU. A TypeScript agent whose local-models section sets url on the model object is How to run Mastra agents with a self-hosted model on a cloud GPU. A ChatAgent whose vLLM example sets ModelPlatformType.VLLM is How to run CAMEL-AI agents with a self-hosted model on a cloud GPU. A local agent UI whose Docker command publishes port 3000 is How to run OpenHands with a self-hosted model on a cloud GPU. The README install is pip install --upgrade metagpt: How to run MetaGPT with a self-hosted model on a cloud GPU. A coding-agent server whose card prints --tool-call-parser qwen3_coder is How to serve Qwen3-Coder for coding agents on a cloud GPU. A general vLLM install is How to serve an OpenAI-compatible API with vLLM on a cloud GPU. A general SGLang install is How to run SGLang on a cloud GPU. Ollama’s install and port 11434 are How to run Ollama on a cloud GPU. CloudGPU’s hosted gateway is a different base URL: How do I use an OpenAI-compatible API for open models?. How published weight sizes are read: How much VRAM do I need to run an LLM?. The tunnel is How to use SSH port forwarding on a cloud GPU. If the process runs out of memory: How to fix CUDA out of memory on a cloud GPU. If the server raises no kernel image on an RTX 5090: How to fix CUDA no kernel image on an RTX 5090. A library whose README prints uv add langchain, and whose vLLM chat page sets ChatOpenAI base_url to http://localhost:8000/v1, is How to build LangChain agents with a self-hosted model on a cloud GPU. A RAG app whose v1.0.0 Compose line is docker compose -f docker-compose.yml up -d, and whose local-model page sets a vLLM base URL to http://host.docker.internal:8000/v1, is How to run RAGFlow with a self-hosted model on a cloud GPU. An OpenAI-compatible engine whose CUDA 12 container tag is localai/localai:latest-gpu-nvidia-cuda-12, and whose quickstart curls http://localhost:8080/v1/chat/completions, is How to run LocalAI on a cloud GPU. A voice framework whose LLM page sets base_url on OpenAILLMService is How to run Pipecat voice agents with a self-hosted model on a cloud GPU. An agent session whose OpenAI-compatible page sets base_url and baseURL on openai.LLM is How to run LiveKit Agents with a self-hosted model on a cloud GPU. A Docker server whose 2.3 quickstart publishes port 80 and whose chat example posts to /v1/chat/completions is How to run GPUStack on a cloud GPU. A framework whose 2.0.11 quickstart pins agentscope==2.0.11, and whose OpenAICredential field base_url is passed to the chat client, is How to run AgentScope agents with a self-hosted model on a cloud GPU. A framework whose README comment sets model_server to http://localhost:8000/v1 and api_key to EMPTY is How to run Qwen-Agent with a self-hosted model on a cloud GPU. A library whose local-LLM page sets OpenAIGPTConfig chat_model and api_base for a vLLM-style endpoint is How to run Langroid agents with a self-hosted model on a cloud GPU. An agent whose Ollama page sets llm to an ollama/ tag, and whose default base URL text is http://localhost:11434/v1, is How to run PraisonAI agents with a self-hosted model on a cloud GPU. An agent whose OllamaPromptDriver sets model to llama3.1, and whose OpenAiChatPromptDriver example sets base_url to http://127.0.0.1:1234/v1, is How to run Griptape agents with a self-hosted model on a cloud GPU. An agent whose local branch sets base_url to http://localhost:11434/v1 and passes that client into AgentConfig is How to run Atomic Agents with a self-hosted model on a cloud GPU. The LICENSE file says content outside the litellm/enterprise directory is available under the MIT license, and the copyright line is Copyright © 2023 Berri AI.

The steps below are copied from the upstream files and docs pages named in each step. They were not run on a CloudGPU instance for this page. Anything this page adds around them is marked. The README and the provider pages name other providers. This page does not list those names. Example model ids that name a closed-model vendor, or that the issue for this page excludes, are not copied.

What the docs print about memory

The README, litellm/proxy/README.md, the Docker quickstart read on 2026-11-14, the Ollama provider page, the vLLM provider page, and the OpenAI-compatible endpoints page print no GPU GB figure. This page does not mark a catalog row as enough. The proxy is a gateway in front of a model server. The server’s memory is a different page: How much VRAM do I need to run an LLM?.

GET /v1/public/compute-capabilities does not list GPUs. It returns feature flags. The memory column is vram_gb from GET /v1/public/gpu-catalog (generated_at 2026-11-14T18:05:24.586Z). The catalog response also includes an availability field and a price field. This page does not copy them. Every catalog row below has min_gpu_count 1. Rows are sorted by vram_gb, then by name. Which listing you can start is on Deploy.

The catalog template list had no LiteLLM template. It did have Ollama templates (jupyterlab-ollama, ollama-qwen3, ollama). Those templates are a different install. This page does not treat them as LiteLLM Proxy.

Catalog name	vram_gb
RTX 2080	8
RTX 3080	10
RTX 2080 Ti	11
RTX 3080 Ti	12
V100 16G	16
A30	24
RTX 3090	24
RTX 4090	24
RTX 4090D	24
RTX 4080 SUPER 32G	32
RTX 5090	32
RTX 5090D	32
V100 32G	32
A100 40G	40
A800 40G	40
A40	48
RTX 4090 48G	48
RTX 4090D 48G	48
RTX A6000	48
Ascend 910B	64
NVIDIA RTX PRO 5000 72G	72
RTX PRO 5000 72G	72
A100 80G	80
A800 80G	80
H800	80
RTX 6000D	83
H20 96G	96
RTX PRO 6000 96G	96
RTX PRO 6000 Max-Q	96
H20-3e 141G	141

Install the proxy

Repository HEAD was ec48e2491c5132951cd2710671506019ce93f0a0, committed 2026-11-14T17:08:03Z. pyproject.toml at that commit sets version to 1.111.0. PyPI on 2026-11-14 reported litellm 1.109.1. Those are not the same number. This page does not pick one.

The README’s proxy install, copied. The --model value on the next line of that README is not copied.

uv tool install 'litellm[proxy]'

pyproject.toml names an optional extra proxy. The README does not print pip install 'litellm[proxy]'. litellm/scripts/quickstart.sh at this commit, in the branch that finds no container engine, prints pip install litellm and says that line is for calling models from Python. That line does not include the proxy extra. A different repository’s notebook prints pip install 'litellm[proxy]': How to run AutoGen agents with a self-hosted model on a cloud GPU. This page does not treat that notebook as the LiteLLM README.

litellm/proxy/README.md prints a different install, copied:

uv tool install litellm

litellm --model ollama/codellama

The comment under that command says INFO: Ollama running on http://0.0.0.0:8000. The root README’s client uses http://0.0.0.0:4000. Those ports are not the same. This page does not merge them.

The Docker quickstart, read 2026-11-14, and the comment header of litellm/deploy/docker/docker-compose.quickstart.yml at this commit, copied. A timing sentence on that page is not copied.

curl -sSLO https://github.com/BerriAI/litellm/raw/main/litellm/deploy/docker/docker-compose.quickstart.yml
printf 'LITELLM_MASTER_KEY=sk-%s\nLITELLM_SALT_KEY=sk-%s\n' "$(openssl rand -hex 32)" "$(openssl rand -hex 32)" > .env
docker compose -f docker-compose.quickstart.yml up -d

The compose file sets image: docker.litellm.ai/berriai/litellm:latest and publishes port 4000. The same page says to pin a tested release tag or digest before production. The quickstart script writes LITELLM_BIND=127.0.0.1: into .env. The compose comment says that value keeps a new install listening on this machine only. litellm/deploy/docker/README.md also prints a checkout path, cd litellm/deploy/compose, then the same printf and docker compose up -d --build. Without --build, that README says docker compose up pulls the published main-stable image. This page does not merge latest and main-stable.

The pages this guide read do not print hf download or huggingface-cli. Weight download for the model server is a different guide: How to download a Hugging Face model.

Point the proxy at the model server

The Ollama provider page’s proxy config that sets api_base, copied:

model_list:
  - model_name: "llama3.1"
    litellm_params:
      model: "ollama/llama3.1"
      api_base: "http://localhost:11434"

Start command on that section, copied:

litellm --config /path/to/config.yaml --detailed_debug

The comment under that command says RUNNING ON http://0.0.0.0:4000.

The same page’s tool-calling config, copied. It does not set api_base.

model_list:
  - model_name: "llama3.1"
    litellm_params:
      model: "ollama_chat/llama3.1"
      keep_alive: "8m" # Optional: Overrides default keep_alive, use -1 for Forever
    model_info:
      supports_function_calling: true

The vLLM provider page says the provider route for an OpenAI-compatible server is hosted_vllm/. Its proxy example sets model_name to my-model, api_base to https://hosted-vllm-api.co, and model to hosted_vllm/ plus a repository id this page does not copy. A later example on that page sets model to hosted_vllm/ plus an id this page does not copy. The embedding section sets HOSTED_VLLM_API_BASE to http://localhost:8000. A vLLM server’s listen address is a different document: How to serve an OpenAI-compatible API with vLLM on a cloud GPU. This page does not merge that guide’s flags into this config.

The OpenAI-compatible endpoints page’s proxy config, copied. The angle-bracket tokens are the placeholders that page prints.

model_list:
  - model_name: my-model
    litellm_params:
      model: openai/<your-model-name>  # add openai/ prefix to route as OpenAI provider
      api_base: <model-api-base>       # add api base for OpenAI compatible provider
      api_key: api-key                 # api key to send your model

That page says if you see Not Found Error when testing, make sure api_base has the /v1 postfix, and its example is http://vllm-endpoint.xyz/v1. The same page’s completion example sets model to openai/mistral and api_base to http://0.0.0.0:4000. A later config on that page names a tag this page does not copy. This page does not copy that config.

Master key, virtual key, and a request

LITELLM_MASTER_KEY is the key the compose file requires. The Docker quickstart says it authorizes every management API call and is the Admin UI password. The username that page prints is admin. The UI URL that page prints is http://localhost:4000/ui.

The virtual keys page, read 2026-11-14, sets the master key in config, copied in part. The model_name values on that block are not copied.

general_settings:
  master_key: os.environ/LITELLM_MASTER_KEY

The same page says you generate keys by hitting /key/generate. The URL, headers, and --data-raw are copied. --data-raw is what makes this a POST. The model names in that page’s JSON are not copied. models here is llama3.1, the model_name the Ollama config above already prints. The metadata user on that page is not copied.

curl 'http://0.0.0.0:4000/key/generate' \
  --header 'Authorization: Bearer <your-master-key>' \
  --header 'Content-Type: application/json' \
  --data-raw '{"models": ["llama3.1"]}'

The page says the call returns a key. It also says virtual keys require a database, and that a request carrying one fails with No connected db. when no database is connected.

The Docker quickstart posts to this URL. The bearer header and the message are copied. The model string on that page is not copied. This block sets model to llama3.1, the model_name the Ollama config above already prints. The OpenAI-compatible config’s model_name is my-model.

curl http://localhost:4000/v1/chat/completions \
  -H 'Authorization: Bearer sk-<your-virtual-key>' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "llama3.1",
    "messages": [{"role": "user", "content": "Say hello in five words."}]
  }'

The Python client on the same page sets base_url to http://localhost:4000 and api_key to that virtual key. The OpenAI-compatible endpoints page’s client, copied. my-model is the model_name from the config above, not a weight id.

import openai
client = openai.OpenAI(
    api_key="sk-<your-litellm-api-key>",
    base_url="http://0.0.0.0:4000"
)
response = client.chat.completions.create(
    model="my-model",
    messages = [
        {
            "role": "user",
            "content": "what llm are you"
        }
    ],
)
print(response)

The simple proxy page says a client that already speaks this request format can send it to the gateway, and that you do not change the client code. An agent points its OpenAI-compatible base URL at the proxy URL those clients already print. None of these pages print an ssh command. How to use SSH port forwarding on a cloud GPU is how a program on your computer reaches port 4000 on the instance. This page does not add an ssh command. After a forward that lands on port 4000 of your computer, the URL the quickstart already prints is http://localhost:4000.

When it fails

Master key missing. litellm/deploy/docker/README.md says the error required variable LITELLM_MASTER_KEY is missing a value means Compose did not find a .env file with LITELLM_MASTER_KEY in the directory you ran it from.

Database password. The same README says password authentication failed for user "llmproxy" when the litellm_postgres_data volume was initialised by an older compose file. It says docker compose down -v drops that volume. This page did not run that command. The quickstart file in this commit names the volume postgres_data and the user litellm. Those strings are not the same as llmproxy. This page does not merge them.

Not Found Error. The OpenAI-compatible endpoints page says to check that api_base has the /v1 postfix.

No connected db. The virtual keys page says a request that carries a virtual key fails with that string when no database is connected. The master key is the key the same page says works without that database. A budget sentence on the Docker quickstart prints an amount. This page does not copy it.

CUDA out of memory. The pages this guide read print no GPU GB figure. If the model server raises CUDA out of memory, use how to fix CUDA out of memory on a cloud GPU. This page does not pick a smaller tag for you.

“CUDA error: no kernel image is available for execution on the device” on an RTX 5090. The pages this guide read do not pin a PyTorch build and do not name compute capability 12.0. The RTX 5090 guide says that card is compute capability 12.0 (sm_120). This page did not re-test a model server on an RTX 5090. Catalog RTX 5090 and RTX 5090D are 32. That is the catalog number only.

What to copy off, then destroy

The quickstart tells you to keep .env, because regenerating LITELLM_SALT_KEY makes credentials already stored in the database unreadable. Copy .env and the config file you passed to --config before you destroy the instance. The Postgres volume named in the compose file is on the instance disk. The model server stops with the instance. Save your work before you destroy a cloud GPU is the checklist. The terms say data on local instance storage is permanently deleted when the instance is destroyed.

The billing page states the billing rule in these words:

Per-minute options have a 1-minute minimum; unused prepaid minutes are refunded when you destroy the instance. Some configurations use whole-hour billing with a 1-hour minimum; unused time within a charged hour is not refunded on voluntary stop. The billing period, minimum charge and initial prepayment are shown before confirmation. No subscription.

Amounts are not copied here.

Frequently asked questions

What memory figure do the LiteLLM proxy pages print?

The README, litellm/proxy/README.md, the Docker quickstart, the Ollama provider page, the vLLM provider page, and the OpenAI-compatible endpoints page print no GPU GB figure. This page does not mark a catalog row as enough. Which listing you can start is on https://cloudgpu.app/deploy.

Which settings point the proxy at a server on the GPU?

The Ollama provider page sets model to ollama/llama3.1 and api_base to http://localhost:11434, with model_name llama3.1. Its tool-calling block sets model to ollama_chat/llama3.1 and supports_function_calling to true, and does not set api_base. The vLLM provider page uses the hosted_vllm/ prefix. Its example repository id is not copied. The OpenAI-compatible endpoints page sets model to openai/ plus a placeholder and api_base to a placeholder, and says the base should end in /v1. Those pages do not print an SSH command. This page did not run them.

How does a client call the proxy?

The Docker quickstart posts to http://localhost:4000/v1/chat/completions with a virtual-key bearer header. The message content it prints is Say hello in five words. The model string on that curl is not copied. This page sets model to llama3.1, the model_name from the Ollama config. The OpenAI-compatible endpoints page sets base_url to http://0.0.0.0:4000 and model to my-model. Virtual keys are created with POST /key/generate. The JSON body sets models to llama3.1. The model names in the upstream generate body are not copied. The Admin UI URL is http://localhost:4000/ui and the username is admin.

What license sentence does the file print?

The LICENSE file says content outside the litellm/enterprise directory is available under the MIT license, and the copyright line is Copyright © 2023 Berri AI. Repository HEAD was ec48e2491c5132951cd2710671506019ce93f0a0, committed 2026-11-14T17:08:03Z. pyproject.toml at that commit sets version to 1.111.0. PyPI on 2026-11-14 reported 1.109.1. That license text is not a CloudGPU rate.

Sources (read 2026-11-14, Asia/Shanghai)

• BerriAI/litellm commit ec48e2491c5132951cd2710671506019ce93f0a0, committed 2026-11-14T17:08:03Z. pyproject.toml version 1.111.0. PyPI litellm 1.109.1 the same day. LICENSE: content outside litellm/enterprise is MIT, Copyright © 2023 Berri AI
• README at that commit. uv tool install 'litellm[proxy]'. Client base URL http://0.0.0.0:4000. The --model value and the other provider names in the completion examples are not copied. A latency figure on that page is not copied
• litellm/proxy/README.md at that commit. uv tool install litellm. litellm --model ollama/codellama. Comment http://0.0.0.0:8000. The client model id on that page is not copied
• litellm/scripts/quickstart.sh at that commit. pip install litellm only in the no-container-engine message. Generated .env includes LITELLM_MASTER_KEY, LITELLM_SALT_KEY, and LITELLM_BIND=127.0.0.1:
• litellm/deploy/docker/docker-compose.quickstart.yml and litellm/deploy/docker/README.md at that commit. Image docker.litellm.ai/berriai/litellm:latest. Port 4000. Volume postgres_data. main-stable when docker compose up is run without --build. Error strings for a missing master key and for user llmproxy
• https://docs.litellm.ai/docs/proxy/docker_quick_start read 2026-11-14. Admin UI http://localhost:4000/ui, username admin. Curl to http://localhost:4000/v1/chat/completions. The page’s model string is not copied. This page sets model to llama3.1, the Ollama config’s model_name. A timing sentence and a budget amount are not copied
• https://docs.litellm.ai/docs/simple_proxy read 2026-11-14. Gateway description. The example model id is not copied
• https://docs.litellm.ai/docs/providers/ollama read 2026-11-14. ollama/llama3.1, api_base http://localhost:11434, ollama_chat/llama3.1, supports_function_calling: true, http://0.0.0.0:4000. Other tags in the model table are not copied
• https://docs.litellm.ai/docs/providers/vllm read 2026-11-14. Prefix hosted_vllm/. api_base https://hosted-vllm-api.co. HOSTED_VLLM_API_BASE http://localhost:8000. Example repository ids are not copied
• https://docs.litellm.ai/docs/providers/openai_compatible read 2026-11-14. Config placeholders. Not Found Error and the /v1 postfix. Client base_url http://0.0.0.0:4000, model my-model. One later config is not copied
• https://docs.litellm.ai/docs/proxy/virtual_keys read 2026-11-14. general_settings.master_key: os.environ/LITELLM_MASTER_KEY. POST /key/generate with --data-raw. The body on this page sets models to llama3.1. The model names, the metadata user, and spend figures on that page are not copied. No connected db.
• The pages above do not print hf download or an ssh command
• Public GPU memory: https://cloudgpu.app/v1/public/gpu-catalog (generated_at 2026-11-14T18:05:24.586Z). GET /v1/public/compute-capabilities returned feature flags and no GPU list. The catalog had no LiteLLM template. It listed Ollama template ids jupyterlab-ollama, ollama-qwen3, and ollama

---

Cards you can start are on the Deploy page. Amounts are not copied here.

Last updated: 2026-11-14 (Asia/Shanghai).