Best AI model for…
Shortlists for common needs, rebuilt from fresh data every day. They rank by price, limits and features. Quality claims only appear where we ran a Hero test.
Cheapest AI models with tool calling
Agents and automations need tool (function) calling. These are the lowest priced models that support it, with at least 100K context.
Currently #1: Qwen3.7 Flash
AI models with the largest context window
How much text fits into a single request? Ranked by context window size, cheaper model first on a tie.
Currently #1: GPT-6 Luna
Free LLM APIs
Language models you can call through an API at no charge right now. Free tiers are rate limited and can disappear without notice.
Currently #1: North Small Translate
Cheapest AI models that can read images
Models that accept images as input (screenshots, photos, scans), ranked by price.
Currently #1: Qwen3.7 Flash
Cheapest reasoning models
Models with a switchable reasoning (thinking) mode, ranked by price. Reasoning tokens are billed as output.
Currently #1: Qwen3.7 Flash
1M-token context for under $1 input
Models that read at least one million tokens per request and cost no more than $1 per million input tokens.
Currently #1: Qwen3.7 Flash
Cheapest models with prompt caching
Prompt caching makes repeated context much cheaper. These models list a cached input price, ranked by standard price.
Currently #1: Qwen3.7 Flash
Newest AI language models
The latest language models in the catalog, newest first.
Currently #1: Gemini 4 Hydrogen
