Hi, I have an existing install and I want to update to use the n8n AI assistant. I was looking through the docs and it says to put the version number of the n8n instance in the .env file. I’m using Docker Compose and currently I do not need to put the version number, which makes it easier to update automatically when there’s a new version. What I’m trying to understand is whether I will now need to manually update the .env file every time I update n8n and put the exact version number to avoid breakage. Thanks, L

Hey, you’re definitely getting closer :grinning_face_with_smiling_eyes: not 100 percent sure to be honest but the new error points to the API - runner authentication missing x-api-key I’d check that these two resolve to the same value sandbox-api: SANDBOX_API_RUNNER_API_KEY: ${SANDBOX_API_RUNNER_API_KEY} sandbox-runner-1: SANDBOX_RUNNER_API_KEYS: ${SANDBOX_API_RUNNER_API_KEY} That’s how I have it in my working setup as well. After checking that recreate both containers rather than only restarting them. The Python warning is separate — Python installed on the Docker host isn’t automatically available inside the n8n container.

Hey @DupervalAI, while you wait for a response, here are some things that might help: Suggested resources Automatically matched to your question. Docs: One-line setup Environment variable installation v2.0 Breaking changes Forum: N8n AI Assistant on Self Hosted Best practices for updating self-hosted n8n across multiple versions? N8n not updating (Docker Compose) @kjooleng, @A_A4, @rantoniuk - you’ve helped with similar issues before, can you take a look? Automatically suggested by n8n’s community bot. It’s a pilot - please share feedback here.

Hi, I jsut had a look at my own self-hosted setup because I’m already running the n8n AI Assistant with the sandbox service using Docker Compose. Also installed it on my existing docker/n8n setup. In my compose file I still use: image: docker.n8n.io/n8nio/n8n:latest and the sandbox images are using latest as well. So you don’t technically need to maintain an N8N_VERSION value in .env just because you want to use the AI Assistant. The version variables in the documentation are used because their example compose file references those variables to pin the images to specific versions. If your existing compose file continues to use :latest, you can keep your existing update approach. I would therefore adapt the Assistant/sandbox parts to your existing compose setup rather than replacing your whole setup with the example from the documentation. I’m running it this way myself and the Assistant + sandbox works fine. Hope that helps! here`s the orig doc i used to install [imagen] Install using Docker Compose | Deploy | n8n Docs Build a Docker Compose setup for self-hosted n8n, including the sandbox stack for n8n Assistant.

Thanks, that’s the file I was looking at also, and the first line there mentioned the n8n version, which is why I came here to ask. My compose file doesn’t specify “latest” anywhere. But every Sunday morning it stops the running instance, does an update of everything, and restarts it. So I’m assuming I’m always on the latest version. I’m guessing the n8n sandbox version might be similar. Thanks, I will proceed and come back here to give the results. L

oi @DupervalAI ! contribuindo com um ponto a mais: Você só precisa atualizar N8N_VERSION manualmente se o seu compose.yml usar essa variável na tag da imagem. Se a sua instalação atual atualiza o n8n automaticamente sem N8N_VERSION, você pode manter esse funcionamento ao adicionar o Assistant. O ponto de atenção são os novos containers de sandbox do Assistant: confirme que sua rotina semanal também baixará as imagens novas e recriará esses containers. Apenas reiniciar não atualiza imagens. Se quiser mais controle e previsibilidade, use versões fixadas para o n8n e para os serviços de sandbox e atualize todas de forma coordenada.

Hi, Since the setup instructions required the version numbers in the configuration, I left it as is and I put my version to “latest”. I just did a stop and restart of my Docker and on restart it seems to be fine. I’m waiting for the next trigger to run in about five minutes to make sure that everything is working. If all of that works, I’m going to stop it again, configure the AI assistant, and then restart and see. I just need to decide if I’m going with one of the frontier models or another model like Kimi/DeepSeek through OpenRouter. Then I’ll do a compose pull to update n8n and all should be good. L

Hi @DupervalAI, One small correction before you continue: N8N_VERSION=latest is a mutable Docker tag, not an exact version number. It can work if you intentionally want every update to follow latest, but the Assistant stack has a separate version variable: N8N_VERSION is shared by the n8n and task-runner images. N8N_SANDBOX_VERSION is shared by the sandbox API, sandbox runner, and sandbox image. The sandbox version is independent of the n8n version. Use the latest stable n8n-sandbox-service release for N8N_SANDBOX_VERSION; do not copy your n8n version into it. The three sandbox images should remain on the same sandbox release. Also, use this update sequence: docker compose pull docker compose up -d docker compose pull only downloads the images. A stop/start or restart keeps the existing containers, while up -d recreates services whose images changed. Since this is an existing installation, keep your current n8n volume, database settings, and encryption-key configuration when adding the sandbox services. Do not replace the whole Compose file blindly, or you may start n8n against a new empty volume. You can check the resolved images with: docker compose config --images Current setup reference: [imagen] Install using Docker Compose | Deploy | n8n Docs Build a Docker Compose setup for self-hosted n8n, including the sandbox stack for n8n Assistant.

Hi, I didn’t blindly change the compose file. I added the pieces that were pertinent. I did them one at a time to make sure that when I was adding a new feature, I did not break something that was working before. My last scheduled run worked fine, so I proceeded to the AI part. I hit a wall because when I tried to configure the AI service in the n8n console, I got a 404 error. I fixed the provider URL to use OpenROuter and now I get: The service couldn't complete the test. Check its status and your settings, then try again. Error details: no sandbox runners are registered or available I’m not sure why. Everything seems to have started correctly. Nevermind, the API runners aren<T listed. NAME IMAGE root-n8n-1 ``docker.n8n.io/n8nio/n8n root-runners-1 ``ghcr.io/n8n-io/runners:latest root-sandbox-api-1 ``ghcr.io/n8n-io/n8n-sandbox-service-api:latest root-searxng-1 ``ghcr.io/searxng/searxng:latest root-traefik-1 traefik I’ll have to dig trhough the logs. L

awwwww :frowning: That error points to the sandbox side rather than OpenRouter no sandbox runners are registered or available Your sandbox API seems to be running. But it doesn’t look like a sandbox runner has successfully registered with it. I’m running the Assistant with the sandbox service as well and my setup has a separate n8n-sandbox-service-runner-dind container connected to the sandbox API I would check the sandbox-api and runner logs next. if you share those we can probably see why the registration is failing…i hope

[imagen] DupervalAI: I hit a wall because when I tried to configure the AI service in the n8n console, I got a 404 error. Make sure you have this # ─── Sandbox ───────────────────── sandbox-certs: image: ghcr.io/n8n-io/n8n-sandbox-service-api:latest container_name: sandbox-certs restart: unless-stopped user: '0:0' entrypoint: ['sh', '-c'] command: - > bootstrap-mtls.sh --out-dir /tls --api-san sandbox-api --control-san-prefix sandbox-runner --world-readable && chown -R sandbox-api:sandbox-api /tls/api && chmod -R a+rX /tls environment: NUM_RUNNERS: '1' volumes: - sandbox-tls:/tls sandbox-api: image: ghcr.io/n8n-io/n8n-sandbox-service-api:latest container_name: sandbox-api restart: unless-stopped depends_on: sandbox-certs: condition: service_completed_successfully environment: SANDBOX_API_GRPC_TLS_CERT_FILE: /tls/api/grpc-server.crt SANDBOX_API_GRPC_TLS_KEY_FILE: /tls/api/grpc-server.key SANDBOX_API_GRPC_TLS_CLIENT_CA_FILE: /tls/api/ca.crt SANDBOX_API_RUNNER_CONTROL_GRPC_TLS_CA_FILE: /tls/api/ca.crt SANDBOX_API_RUNNER_CONTROL_GRPC_TLS_CERT_FILE: /tls/api/control-grpc-api-client.crt SANDBOX_API_RUNNER_CONTROL_GRPC_TLS_KEY_FILE: /tls/api/control-grpc-api-client.key SANDBOX_API_RUNNER_CONTROL_GRPC_TLS_SERVER_NAME: sandbox-runner-1 volumes: - sandbox-tls:/tls:ro healthcheck: test: ['CMD', 'wget', '-qO-', 'http://localhost:8080/healthz'] interval: 5s timeout: 3s retries: 5 start_period: 10s # Never publish 8080/9090 to the host on an internet-facing server. # n8n reaches this container by service name, over the default Compose network. sandbox-runner-1: image: ghcr.io/n8n-io/n8n-sandbox-service-runner-dind:latest container_name: sandbox-runner restart: unless-stopped privileged: true depends_on: sandbox-api: condition: service_healthy environment: SANDBOX_RUNNER_API_GRPC_ADDR: sandbox-api:9090 SANDBOX_RUNNER_HTTP_BASE_URL: http://sandbox-runner-1:8080 SANDBOX_RUNNER_CONTROL_GRPC_LISTEN_ADDR: ':9091' SANDBOX_RUNNER_CONTROL_GRPC_ADVERTISE_ADDR: sandbox-runner-1:9091 SANDBOX_RUNNER_ID: runner-1 SANDBOX_RUNNER_DOCKER_SANDBOX_IMAGE: ghcr.io/n8n-io/n8n-sandbox-service-sandbox:latest SANDBOX_RUNNER_REGISTRATION_GRPC_CA_FILE: /tls/runner/ca.crt SANDBOX_RUNNER_REGISTRATION_GRPC_CERT_FILE: /tls/runner/grpc-client.crt SANDBOX_RUNNER_REGISTRATION_GRPC_KEY_FILE: /tls/runner/grpc-client.key SANDBOX_RUNNER_REGISTRATION_GRPC_SERVER_NAME: sandbox-api SANDBOX_RUNNER_CONTROL_GRPC_TLS_CERT_FILE: /tls/runner/control-grpc-server.crt SANDBOX_RUNNER_CONTROL_GRPC_TLS_KEY_FILE: /tls/runner/control-grpc-server.key SANDBOX_RUNNER_CONTROL_GRPC_TLS_CLIENT_CA_FILE: /tls/runner/ca.crt volumes: - sandbox-tls:/tls:ro # Never expose this container's ports publicly — it runs privileged Docker-in-Docker. What does docker ps show? I believe your sandbox runner container and image is missing

You may want to check how the AI Assistant setup handles the pinned n8n version alongside your Docker Compose update strategy. I’d avoid changing your automatic update setup until you confirm the versioning requirement, since the exact behavior can depend on how the environment variables are configured.

Asistente N8n, .env e instalaciones existentes

Preguntas

n8n 18 Septiembre, 2026 17:13 2

Hola @DupervalAI, mientras esperas una respuesta, aquí hay algunas cosas que podrían ayudarte:

Recursos sugeridos

Coincidencia automática con tu pregunta.

Documentación:

Foro:

@kjooleng, @A_A4, @rantoniuk - ya han ayudado con problemas similares antes, ¿pueden echar un vistazo?

Sugerencia automática del bot de la comunidad de n8n. Es un piloto - comparte tu opinión aquí.

Tema		Respuestas	Vistas
How can i get the service URL and the API key to enable the sandbox ai assistance, because it is compusory to have it Questions deployment , bug:unconfirmed	22	696	30 Septiembre 2026
How to upgrade n8n with docker-compose setup? Questions deployment , docker , docker-compose	57	26028	2 Abril 2025
I am running Docker in my desktop to run N8N workflow and want to enable the AI Assistance, but have to enter the Service URL and API key. When i go into the .env file, The following key and API is display Nodes	1	91	18 Septiembre 2026
New Self-Hosting n8n Locally + AI Assistant Setup Community Highlights featured	0	1504	25 Agosto 2026
Upgrading to N8N AI Questions docker-compose , langchain	7	1958	12 Octubre 2023

Asistente N8n, .env e instalaciones existentes

Temas relacionados