Research Engine
Research the AI ecosystem through verified relationships.
Start from a category, follow products to models and APIs, trace providers, then open comparisons and sources. The views below are generated from the same structured graph used across iilib.
Filter models by context, price and API support →
Start with your actual constraints instead of browsing every model record.
Follow tools and agents to models and APIs →
See documented compatibility routes before choosing a coding stack.
Practical decision guides
Choose a model or coding agent by the work you need done
Shortlists, costs and concrete testing criteria, backed by the research records below.
GPT-6 Luna vs Sol vs Astra
Pick an API model based on task difficulty and token costs, with a clear example and source links.
OpenCode vs Cline vs Aider
Compare interface, Git workflow, model access, approvals and total usage cost.
Reference views
Generated research assets
These views are generated from the same entity and relationship graph used by the rest of iilib.
AI Coding Landscape
Compare the currently mapped coding tools and agents by provider, product surface, open-source status, model/API links and published comparisons.
LLM API Pricing
Compare published token prices, context windows, providers and API availability from the model graph.
Model Context Windows
Compare context limits, maximum output, modalities, providers, families and API availability.
Open Models Research
Filter representative open-weight models by license, architecture, use case, scale, context, self-hostability, transparency and API availability.
AI Company → Product Graph
Trace companies to directly mapped model families, models, APIs, tools and agents.
Research paths
Start from a mapped question space
Category views are generated from direct membership plus verified graph connections. They are not keyword landing pages.
AI Code Editors & IDEs
AI-enabled code editors and integrated development environments with built-in agent workflows for working on software projects.
AI Coding
Tools and coding agents used to write, edit, review, debug, and maintain software with language models.
LLM APIs
APIs that provide programmatic access to language models for generation, tool use, coding agents, structured output, and related model workflows.
Open-source Coding Agents
Open-source coding agents and developer tools whose client or core agent implementation is available in a public source repository.
Terminal Coding Agents
Coding tools and agents that can inspect repositories, edit files, and carry out development work from a terminal or command-line interface.
Provider graph
Organizations with the most mapped entities
Counts include indexed models, model families, APIs, tools and agents connected through develops, provides, operates or owns relationships.
| Provider | Mapped entities |
|---|---|
| OpenAI | 17 |
| Alibaba Cloud | 9 |
| Anthropic | 7 |
| 7 | |
| DeepSeek | 6 |
| Ai2 | 3 |
| Cognition | 3 |
| Meta | 3 |
| Mistral AI | 3 |
| Z.ai | 3 |
| Amazon Web Services | 2 |
| SpaceXAI | 2 |
Compare
Move from graph context to direct comparison
Published comparisons combine structured entity facts, provider and category context, model/API links, curated differences and source evidence.
Browse 30 comparisons →Evidence
Check the source layer
110 of 110 indexed factual entities currently reference at least one primary source. Relationship evidence is tracked separately and merged into research views where relevant.
Read the methodology →