EUGDPR-compliant models for your agent

Europe-hosted AI Models for
OpenClaw & Hermes Agent

Your agent stays the same — LLMBase supplies Deepseek, Claude, Gemini, and more. EU servers, GDPR-compliant, with the Pro budget from €19/month.

Get startedDocumentation

Fair-use agent budget — pay-per-use stays separate

Integration

Up and running in five minutes.

Your agent stays exactly as-is. Point it at LLMBase instead of OpenAI — and you get EU servers, a wider model choice, and no per-token anxiety.

Agent docs Inference API (pay-per-use) 
# OpenClaw → Settings → Providers → Add custom provider

Provider type:  OpenAI-compatible
Name:           LLMBase
Base URL:       https://llmbase.ai/api/v1/agents
API Key:        llmbase_chat_...

# Then pick a model from the LLMBase model catalog, e.g. ~z-ai/glm-latest

Billing

Flat rate or pay-per-use.
Same models either way.

Pick the billing that fits how you work. There's no wrong answer.

Pro Subscription

Recommended
€19/month

Or €190/year

  • Works in OpenClaw, Hermes, and any OpenAI-compatible agent
  • Predictable cost — no per-token surprises
  • Access to Claude, Deepseek, Gemini, and more
  • Also powers LLMBase Chat — one subscription, both
  • Agents stop with a clear budget error when quota runs out
Get Pro — €19/month

Cancel anytime · DPA included

Pay-per-use Inference API

$0.20/1M tokens

Starting from — varies by model

  • Pay only for what you actually use — no subscription
  • Great for high volume, CI pipelines, or batch jobs
  • Base URL: api.llmbase.ai/v1
  • Load credits once, no credit card required
  • Exact token billing per request
View Inference API

Available models

Deepseek, Claude, Gemini.
The right model for every task.

Every model accessible from your agent. Switch models mid-workflow — no reconfiguration needed.

ModelIncluded
~z-ai

GLM 5.2

~z-ai/glm-latest

Included
z-ai

GLM 5.2

z-ai/glm-5.2

Included
~moonshotai

Kimi K2.7 Code

~moonshotai/kimi-latest

Included
moonshotai

Kimi K2.7 Code

moonshotai/kimi-k2.7-code

Included
nvidia

Nemotron 3 Ultra 550B A55B

nvidia/nemotron-3-ultra-550b-a55b

Included
~minimax

MiniMax M3

~minimax/minimax-latest

Included
minimax

MiniMax M3

minimax/minimax-m3

Included
~deepseek

DeepSeek V4 Pro

~deepseek/deepseek-latest

Included
deepseek

DeepSeek V4 Pro

deepseek/deepseek-v4-pro

Included
deepseek

DeepSeek V4 Flash

deepseek/deepseek-v4-flash

Included
~qwen

Qwen3.6-35B-A3B

~qwen/qwen-small-latest

Included
qwen

Qwen3.6-35B-A3B

qwen/qwen3.6-35b-a3b

Included
moonshotai

Kimi K2.6

moonshotai/kimi-k2.6

Included
google

Gemma 4 26B A4B

google/gemma-4-26b-a4b-it

Included
google

Gemma 4 31B

google/gemma-4-31b-it

Included
qwen

Qwen3.5-9B

qwen/qwen3.5-9b

Included
qwen

Qwen3.5-122B-A10B

qwen/qwen3.5-122b-a10b

Included
~qwen

Qwen3.5 397B A17B

~qwen/qwen-latest

Included
qwen

Qwen3.5 397B A17B

qwen/qwen3.5-397b-a17b

Included
black-forest-labs

FLUX 2 Klein

black-forest-labs/FLUX-2-klein-9b

Included
~qwen

Qwen3-VL-30B-A3B-Instruct

~qwen/qwen-vl-latest

Included
qwen

Qwen3-VL-30B-A3B-Instruct

qwen/qwen3-vl-30b-a3b-instruct

Included
openai

gpt-oss-20b

openai/gpt-oss-20b

Included
openai

gpt-oss-120b

openai/gpt-oss-120b

Included
~qwen

Qwen3 Coder 480B A35B

~qwen/qwen-coder-latest

Included
qwen

Qwen3 Coder 480B A35B

qwen/qwen3-coder

Included
meta-llama

Llama 4 Maverick

meta-llama/llama-4-maverick

Included
mistral

Mistral Nemo

mistral/mistral-nemo

Included
EUAll models hosted exclusively on EU servers.
Agent docs

Frequently Asked Questions

Does my agent keep working the same way?
Yes. You only change the base URL and API key in your OpenClaw or Hermes config. Your workflow, prompts, and tools stay exactly as they are.
When does Pro make sense vs. pay-per-use?
Pro makes sense if you already use LLMBase Chat daily — you get agents on top for a flat fee and always know your monthly cost. Pay-per-use wins for pure API usage, high volume, or when you need exact per-token accounting.
Which key do I give to OpenClaw or Hermes?
Create a chat agent key (llmbase_chat_...) in the dashboard and paste it into OpenClaw or Hermes. Your account login and inference API keys stay private and separate.
What happens when the included budget is used?
Your agent receives a budget error and pauses. For separate pay-per-use API billing, use a normal Inference API key.

Your agent. Better models. Today.

One config change in OpenClaw or Hermes — and your agent runs on EU-hosted Deepseek, Claude, and Gemini.

Get startedDocumentation

Pro €19/month · Pay-per-use from $0.20/1M tokens · No training · GDPR compliant