Neura Market
ollama logo

ollama

Free

Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Model APIsFreeFree tier
Type
Open Source
Visit Website GitHub (170.6k stars)

About ollama

Ollama is a free, open-source platform that makes it easy to run open models locally on your hardware or scale to the cloud. With a single command, you can set up models like OpenClaw, Claude Code, and thousands more. Ollama offers CLI, API, and desktop apps, along with 40,000+ community integrations. It supports unlimited public models, private model sharing, and keeps your data private—never training on your data. Cloud options provide access to larger models with parallel requests and web information retrieval. Pricing tiers (Free, Pro at $20/mo, Max at $100/mo) accommodate light to heavy usage, with concurrency and usage limits based on GPU time.

Key Features

Run open models locally or in the cloud
Single-command installation (curl or PowerShell)
CLI, API, and desktop application interfaces
40,000+ community integrations
Unlimited public models available
Private model upload and sharing (Pro and Max tiers)
Cloud models hosted in United States, Europe, and Singapore
Concurrent model execution (up to 10 on Max tier)
Usage-based cloud credits that reset every 5 hours and 7 days
Priority support and dedicated Slack channel (Team tier)

Pros & Cons

Pros
  • Free and open source with no cost to start
  • Easy one-command setup across platforms
  • Strong privacy: data never trained on and can run entirely offline
  • Extensive community library of models and integrations
  • Flexible deployment from local hardware to cloud scaling
  • Supports tool calling for real agent workflows
Cons
  • Free tier has limited cloud usage (light usage only, 1 concurrent model)
  • Usage measured by GPU time, not tokens, which can be less predictable
  • Cloud model availability and performance depend on hardware and model size
  • Some advanced features (SSO, MDM installer) only in upcoming Team plan

Best For

Coding automation and AI-assisted developmentDocument analysis and researchChatting with language modelsEvaluating and comparing open modelsRunning AI agents and continuous agent tasksDeep research with large modelsPrototyping and experimenting with open-weight models

Alternatives to ollama

AutoGPT

AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.

aider

aider is AI pair programming in your terminal

axolotl

Go ahead and axolotl questions

Chroma

Open-source embedding database

awesome-claude-code

A curated list of awesome skills, hooks, slash-commands, agent orchestrators, applications, and plugins for Claude Code by Anthropic

agents-course

This repository contains the Hugging Face Agents Course.

FAQ

Which models are available in the cloud?
The full list of cloud-enabled models is available on Ollama's website. Models include popular open models like gpt-oss, deepseek, and others.
Do cloud models support tool calling?
Yes. Cloud models that are trained to support tools are tested for tool calling and real agent workflows before going live.
How is usage measured?
Usage reflects actual utilization of Ollama's cloud infrastructure—primarily GPU time, which depends on model size and request duration. This is different from fixed token or request-based plans.
How many cloud models can I run at once?
Free plan allows 1 concurrent model, Pro allows 3, Max allows 10. Requests beyond the limit are queued.
How does extra usage work?
Pro and Max users can add extra usage balance. Ollama uses your included plan limits first, then draws from extra balance when over limits.