Fieldbook to AIThe plain-English guide to which AI to use

By task

Which AI is best for coding?

Our editorial best pick: Claude Opus 5.5; budget pick: Claude Sonnet 5.5; free option: Claude Free.

Last updated .

Costs here use the prices AI companies charge software developers, so they show relative cost, not your monthly bill. See what ChatGPT, Claude and Gemini charge you.

Picks are editorial judgments from the makers' descriptions, price and our smartness rating. We have not run our own tests yet.

Best pick

Claude Opus 5.5

The highest smartness rating we track (5 of 5), and the maker describes it as strong on large coding projects.

Made by Anthropic, the company behind Claude.
Smartness: ★★★★★ (5 of 5)
Reads at once: about 1,500 pages
1,000-word essay: $0.03 (about 3 cents)
50-page PDF summary: $0.15 (about 15 cents)

To use it in an app: Claude plans and prices.

Budget pick

Claude Sonnet 5.5

Rated 5 out of 5 at about half the price of Claude Opus 5.5.

Made by Anthropic, the company behind Claude.
Smartness: ★★★★★ (5 of 5)
Reads at once: about 1,500 pages
1,000-word essay: $0.01 (about 1 cent)
50-page PDF summary: $0.07 (about 7 cents)

To use it in an app: Claude plans and prices.

Free option

Claude Free

Chat, web search, file creation, memory and the Sonnet and Haiku models. Usage limits reset every 5 hours.

Limited; 'there's no fixed message count' (Anthropic). Up to 5 projects.
Checked 2026-10-16; free plans can change. See every app's free plan.

What a 100-message chat costs with each pick

The best pick costs $5.25; the budget pick costs $2.62. These are developer prices, so they show relative cost, not your monthly bill.

Bar scale
Show the data as a table
ModelCost of a 100-message chat
Claude Sonnet 5.5 (budget pick)$2.62
Claude Opus 5.5 (best pick)$5.25

Bars use a straight scale: a bar twice as long means twice the cost. Only the two picks above. Source: OpenRouter list prices and the assumptions on How we work.

All tasks · All models · How costs are worked out