AI Model Comparison Table
Compare today's leading AI models side by side — context window, modalities, price per million tokens, and what each is best for. Filter, sort, pin to compare, and estimate your monthly cost.
| Pin | Model ↕ | Context ↕ | Modalities | In $/1M ↕ | Out $/1M ↕ | Best for | Open |
|---|
Estimate my monthly cost (optional)
Pin models below to see each one's estimated monthly cost using these volumes.
Side-by-side comparison
We only send the comparison if you ask. No spam.
LLM Context Window Comparison: See Context, Price, and Capability Side by Side
Choosing a model usually means opening eight tabs, each presenting its numbers differently, and trying to hold three context windows in your head while comparing prices quoted per different units. By the time you have the picture assembled, another model has launched.
This table collapses that into one view. Every model sits in the same grid, measured on the same columns, so a like-for-like LLM context window comparison takes seconds instead of an afternoon.
What Each Row Tells You
The table shows the attributes that actually drive a decision:
- Context — how much the model can hold at once, sortable so you can rank by capacity directly.
- Modalities — what kinds of input it accepts.
- Input price per million tokens and output price per million tokens, listed separately because the gap between them changes your bill significantly.
- Best for — the practical positioning, in plain language.
- Open — whether the weights are open.
Sorting is available on model, context, and both price columns, so you can order the whole table by whichever factor matters most to your project.
Filter Before You Compare
Filters narrow the field before you start reading closely. Set a minimum context — any, 32K+, 128K+, 200K+, or 1M+ — to hide anything that cannot handle your documents. Set a maximum input price to remove options outside budget. Filter by modalities if you need more than text. There is also a search box for jumping straight to a named model.
An open-weights only toggle answers a specific question people arrive with: an open source LLM comparison, filtered to models you can self-host or inspect rather than only call through an API. It is one switch instead of a research project.
Pin the Shortlist, Then Estimate the Bill
Every row has a pin. Pinning two or three models isolates your shortlist so the rest of the table stops competing for attention.
Beneath the table sits an optional cost estimator. Enter your input tokens per request, output tokens per request, and requests per month, and each pinned model returns an estimated monthly cost at your actual volumes. This is the step most comparisons omit — a price per million tokens means very little until it meets your traffic.
A Realistic Example: A Document Summarisation Feature
A product team is adding summarisation for uploaded contracts, often forty pages long. Their real constraint is context length; their second constraint is monthly spend.
They set the minimum context filter to 200K+, which immediately removes most of the table. From what remains, they sort by input price ascending and pin the three cheapest that still fit. Then they enter their volumes — roughly 60,000 input tokens per request, 1,500 output, 4,000 requests a month.
The estimator shows the three monthly figures side by side, and the cheapest model that meets the context requirement becomes obvious. The whole evaluation takes ten minutes.
Answering “Which LLM Is the Best” Honestly
There is no universal answer, and any page claiming one is selling something. The best model for long-document work is rarely the best for high-volume classification, and the cheapest LLM by input price may not be cheapest once output-heavy workloads are counted.
The table takes the opposite approach: it presents the specifications and lets you weigh them against your own constraints. If you are evaluating the best open source LLM for coding in 2026, filter to open weights, check the “best for” column, and compare pricing on the same screen — then judge against your own tasks rather than a leaderboard.
Who It Helps
Developers picking a model for a build. Product managers comparing options before committing. Founders weighing capability against burn rate. Anyone who has to justify a model choice to someone else. Once you have picked one, the LLM Token Cost Calculator and API Cost Calculator refine the budgeting.
Why output tokens drive your bill
Every request has two parts: the tokens you send in (your prompt and documents) and the tokens the model writes back. Output tokens usually cost several times more than input tokens.
Drag the slider to see how the same 1 million tokens costs more as the model writes a bigger share of them.
How big is a context window?
Context is how much a model can read at once. As a rough guide, 1,000 tokens is about 750 English words, or around one and a half pages.
Estimates assume about 500 words per page. Set Min context in the table to hide models that cannot hold your documents.
Match the model to the job
Start from what you are building, not from a leaderboard. Pick a job to see where to start in the table.
Compare Before You Commit
Model choice sets your ceiling on capability and your floor on cost. Run an LLM context window comparison above, filter to what your project actually needs, pin your shortlist, and see the monthly figure before you write a line of integration code.
Compare modelsFive questions before you commit
Answer these and the right shortlist usually picks itself. Each one maps to a column or filter in the table.
Open the table- How much does it need to read at once? Set Min context.
- How many requests a month? Put it in the cost estimator.
- Does it write long answers? Weigh the output price.
- Do you need to self-host? Turn on Open-weights only.
- What is your monthly ceiling? Set Max input price.