Tool
Ollama for Hermes Agent: Local Model Server Setup
Use Ollama as Hermes Agent’s local model server with the current one-command launch, manual endpoint, context checks, tool-call tests, and privacy boundaries.
Quick answer
Run `ollama launch hermes` for the fastest setup, or connect Hermes manually to `http://127.0.0.1:11434/v1` through `hermes model`. Ollama serves the model; Hermes adds tools, memory, skills, cron, browser automation, and gateways. Verify a real tool call before relying on the model for unattended work.
Ollama and Hermes solve different layers. Ollama downloads and serves local or cloud models. Hermes is the agent runtime that plans work, calls tools, remembers context, schedules jobs, and connects messaging platforms. Use the [full local LLM guide](/features/local-llm-support) to compare Ollama with LM Studio, vLLM, SGLang, and llama.cpp.
Setup steps
- 1Install Ollama and verify `http://127.0.0.1:11434/api/tags`.
- 2Run `ollama launch hermes` for automatic onboarding, or use `hermes model` for manual endpoint control.
- 3Choose a tool-capable model that fits your hardware and confirm the exact detected model ID.
- 4Run a file-and-terminal task, then test long context and latency before connecting a gateway.
- 5Add a hosted fallback only for workloads allowed to leave the machine.
Command map
ollama launch hermesFastest current path: choose a model and let Ollama configure Hermes.
curl http://127.0.0.1:11434/api/tagsProve the local Ollama server is reachable before debugging Hermes.
hermes modelSelect the local/custom endpoint and exact model manually.
hermes chat -q "Read README.md and report the project name"Acceptance test for real file/tool use, not just conversational output.
Self-hosted Web UI or FlyHermes?
Choose Ollama + self-hosted Hermes
You want model files and inference on hardware you control, accept local GPU/CPU limits, and are willing to own updates, remote access, gateways, backups, and fallback policy.
Avoid when: You need frontier reliability, mobile access, and connected-channel uptime without operating the stack.
Choose FlyHermes
The desired outcome is a managed Hermes experience across browser, phone, and channels without maintaining a model server, VPS, provider routes, or gateway uptime.
Avoid when: Air-gapped inference or direct ownership of the model-serving layer is mandatory.
Features
- ✓One-command `ollama launch hermes` onboarding
- ✓Local OpenAI-compatible endpoint
- ✓Local and Ollama Cloud model selection
- ✓Tool-call and context testing
- ✓Optional messaging gateway setup
Why this tool matters
Local inference removes the two biggest self-host pain points at once — token bills and data exposure. With Ollama, the model runs on your machine, so there are no API charges and no conversation leaves your network.
Hermes requires at least 64,000 tokens of context to maintain working memory for multi-step tool calling, and it rejects smaller windows at startup. For Ollama that means starting the model with -c 65536; this is the single most common reason a local model 'doesn't work' with Hermes.
Local models trade cost for capability. Strong agentic behavior needs a capable model, and the best agentic models are large — so a fully local setup depends on your hardware. Many users run a hybrid: a cheap or local model for routine work and a premium hosted model for hard tasks.
Ollama also has a managed Cloud option (set OLLAMA_API_KEY) if you want Ollama-hosted models without local hardware. That is a middle ground between fully local and a general cloud provider.
Best use cases
FAQ
Run `ollama launch hermes`. Ollama can configure the local endpoint, model, and optional messaging onboarding.
Use the OpenAI-compatible base URL `http://127.0.0.1:11434/v1` on the same machine.
No. Ollama serves models; Hermes provides the agent runtime, memory, tools, skills, cron, browser automation, and gateways.
It keeps model inference local, but network tools, remote MCP, hosted memory, speech providers, and fallbacks can still send data externally.