MODEL ATLAS / OPENROUTER

OpenAI & Anthropic, in perspective.

Find the right model for your budget, with the wider market for context.

Updated Nov 20, 2026 · 22:26 UTC Fresh snapshot
01 / THE SHORTLIST

Best models for your budget

OpenAI & Anthropic · highest Intelligence Index in each relative price group.

Only eligible OpenAI and Anthropic models in the weekly top 30 are considered. They are sorted by blended cost and split into three near-equal groups. Ties favor lower cost, then weekly rank. Groups are relative to this sample, not fixed budgets or guarantees.

LOW COST

Anthropic: Claude Haiku 5.5

anthropic/claude-haiku-5.5
INTELLIGENCE36.5
BLENDED / 1M$0.200
Group range $0.200 – $0.450 / 1M
MEDIUM COST

Anthropic: Claude Sonnet 5.5

anthropic/claude-sonnet-5.5
INTELLIGENCE54.5
BLENDED / 1M$4.000
Group range $2.000 – $4.000 / 1M
HIGH COST

Anthropic: Claude Opus 5.5

anthropic/claude-opus-5.5
INTELLIGENCE59.1
BLENDED / 1M$8.000
Group range $4.000 – $8.000 / 1M
02 / THE LANDSCAPE

Quality vs. cost

Find the frontier: more intelligence for less spend.

3 input : 1 output

Blended USD per million tokens is an illustrative 3:1 input/output workload, not a provider quote. Intelligence Index is one quality proxy, not a universal measure. Models without both prices and a score are omitted.

Showing OpenAI, Anthropic and up to three other publishers with the most eligible models. Select a publisher or a point to highlight it.

No model selected. Select a point to see its details.

Point colors identify the model publisher. The frontier compares only the publishers shown: no other displayed model is both at least as intelligent and no more expensive.
03 / THE DETAILS

Compare the field

Different lenses on the same weekly top 30.

Intelligence Index

Highest score first · models without scores omitted

Input & output pricing

Highest available price first · USD per million tokens · unavailable prices omitted