Free
$0
- Unlimited local use, starter usage credits for starter cloud models, 1 concurrent cloud request
What Is Ollama? Features, Pricing, Pros & Cons
Free and open source for local use, paid plans with monthly cloud usage credits, Enterprise custom
What is Ollama?
Ollama is an open-source tool from Ollama Inc. for running open models like Gemma, Qwen models, DeepSeek, and gpt-oss on your own machine through a CLI and a local REST API. It's MIT-licensed, installs on macOS, Windows, and Linux or runs in Docker, and uses llama.cpp as a supported backend.
Ollama also runs models in its own cloud. Free accounts get starter usage credits, paid plans add monthly credits and higher concurrency, and running models on your own hardware stays unlimited on every plan. Ollama states that prompts and responses are never logged or trained on, and the ollama launch command connects models directly to coding agents like Codex, OpenCode, and Claude Code, which makes it easy to run open models on either side of an OpenCode vs Claude Code comparison.
Ollama is built to run models rather than operate them in production. It doesn't route across providers or score outputs, and observability comes from third-party tools that integrate with it.
Key features
Core capabilities this platform advertises.
- ollama run to pull and chat with open models by name
- Local REST API with official Python and JavaScript libraries
- Cloud models billed through usage credits
Strengths and tradeoffs
What this tool does well, and the limitations to keep in mind.
Pros
- MIT-licensed and free to run locally with no usage limits
- One-command install on macOS, Windows, and Linux
- Large library of open models pulled by name
Cons
- Cloud concurrency is capped by plan: 1 request on Free, 3 on Pro, 10 on Max and Team
- No routing across providers or automatic failover
Plans & pricing
What's included in each plan, and how the tiers compare.
Pro
$20/mo or $200/yr
Per month or per year
- $60 of usage credits per month, larger pro models, 3 concurrent requests
Max
$100
Per month
Common use cases
Developers who want to run open models locally for prototyping, private workloads, or local coding agents
- Running open models privately on a laptop or workstation
- Prototyping LLM apps without per-token API costs
- Powering local coding agents
Using Ollama with Respan
Respan has a native Ollama integration. The respan-instrumentation-ollama package patches the official Ollama Python client and sends chat, generation, tool call, and embedding spans to Respan, so you keep running models locally and get tracing, evals, and cost tracking on every call.
- Trace every Ollama chat, generation, and embedding call
- Attach customer IDs, threads, and metadata to local model calls
- Score local model output with the same evals you use in production
- Datasets built from real local runs
- Route to 1,000+ hosted models with automatic fallbacks when you outgrow local
Trace and evaluate your Ollama models with Respan
10KFree traces/mo
1,000+Models
Best Ollama alternatives & competitors
Top companies in Inference & Compute you can use instead of Ollama.
Compare Ollama
Side-by-side comparisons with other tools in this category.
Best integrations for Ollama
Companies from adjacent layers in the AI stack that work well with Ollama.
