Claude Opus 4.7

ModelActive

by Anthropic · family “claude-opus”

Opus 4.7 is the next generation of Anthropic's Opus family, built for long-running, asynchronous agents. Building on the coding and agentic strengths of Opus 4.6, it delivers stronger performance on... (description from the OpenRouter listing)

in: textin: imagein: fileout: textReasoningTool useStructured output

Epoch Capabilities Index

156.4 #19 of 268

90% CI 154.4 – 158.5

Listing ↗

Released

16 Apr 2026

Context window

1M tokens

Max output

128K tokens

Input $ / 1M tokens

$5.00

Output $ / 1M tokens

$25.0

Cached input $ / 1M

$0.50

Each benchmark shown separately with its own source

Benchmark results (19)

All benchmarks

Benchmark	Domain	Score	vs best recorded	Setting	Run	Source
Furniture Assembly	multimodal	33.3% ±6.1	40%	max	10 Sep 2026	Epoch ↗
LMCA	agents	52.2%	80%	max	—	External ↗
DTBench	reasoning	94.7%	96%	max	—	External ↗
Mystery Game Puzzles	games	28.0% ±4.5	33%	max	25 Jul 2026	Epoch ↗
EBR-bench	reasoning	19.0%	25%	max	30 Jun 2026	Epoch ↗
MirrorCode	coding	31.1% ±9.0	42%	high	12 Aug 2026	Epoch ↗
OSWorld 2.0	agents	18.2%	58%	max	—	External ↗
FrontierMath-Tiers-1-3-v2-Private	math	70.2% ±2.7	75%	max	10 Jun 2026	Eval log ↗
FrontierMath-Tier-4-v2-Private	math	31.7% ±7.4	32%	max	10 Jun 2026	Eval log ↗
ProofBench	math	54.0%	54%	max	—	External ↗
APEX-Agents	agents	49.2%	72%	max	—	External ↗
Chess Puzzles	games	7.0% ±2.6	10%	max	6 Aug 2026	Eval log ↗
GSO-Bench	coding	44.1%	94%	high	—	External ↗
ARC-AGI-2	reasoning	75.8%	80%	max	—	External ↗
WeirdML	coding	76.4% ±0.0	82%	high	—	External ↗
OTIS Mock AIME 2024-2025	math	86.7% ±5.1	87%	max	6 Aug 2026	Eval log ↗
SWE-Bench verified	coding	83.5% ±1.7	100%	max	20 Apr 2026	Eval log ↗
GPQA diamond	science	86.4% ±2.4	90%	max	6 Aug 2026	Eval log ↗
ARC-AGI	reasoning	93.5%	95%	high	—	External ↗

Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.

API list price over time

Price history

$ per 1M tokens

6 observations (OpenRouter listing + Internet Archive snapshots). History accumulates with every ingest run; a single point means no change has been observed yet.