← AI news, last 7 days

Claude Opus 5.5

A model by Anthropic · released 2026-09-22 · 151 checked facts from 26 sources

Timeline

Prices

Per price, each value with the dates it was seen. A value seen on one date only is one copy of the page. No price has changed over the record.
Price forValues over time
per 1M cached input tokens$0.20 2026-09-26 to 2026-10-25 · Zvi Mowshowitz, Anthropic's page as archived 2026-10-01, minimaxir.com
per 1M cached input tokens (write, standard)$5 2026-09-22 to 2026-10-01 · Anthropic the lab's own, Anthropic's page as archived 2026-10-01
per 1M input tokens$4 2026-09-22 to 2026-10-25 · Simon Willison, Latent Space, Zvi Mowshowitz, Anthropic's page as archived 2026-10-01, minimaxir.com
per 1M input tokens$4 2026-09-22 · artificialanalysis.ai
per 1M input tokens (fast mode)$8 2026-09-22 to 2026-10-16 · Anthropic the lab's own, Zvi Mowshowitz, Latent Space
per 1M output tokens$20 2026-09-22 to 2026-10-25 · Simon Willison, Latent Space, Zvi Mowshowitz, Anthropic's page as archived 2026-10-01, minimaxir.com
per 1M output tokens$20 2026-09-22 · artificialanalysis.ai
per 1M output tokens (fast mode)$40 2026-09-22 to 2026-10-16 · Anthropic the lab's own, Zvi Mowshowitz, Latent Space
per rollout$1.83 2026-11-03 · Hugging Face the lab's own
per task$12.51 2026-10-30 · Latent Space
unstated (fast mode)$8/$40 2026-10-22 · Zvi Mowshowitz

Scores

Self-reported: stated on the lab's own channel. A score reported twice with two values is shown twice, with its date: a lab re-runs a benchmark under a new setup; neither replaces the other.
BenchmarkScoreDateWho
36 small verification tasks50%2026-11-20reported by Latent Space · Latent Space
AA Coding Agent Index v2.061.72026-10-23reported by Latent Space · Latent Space
AA Coding Agent Index v2.1582026-11-14reported by Latent Space · Latent Space
AA-Omniscience v1.1482026-10-29reported by artificialanalysis.ai · artificialanalysis.ai
APEX-1 + APEX-Agents73.3%2026-10-16reported by mercor.com · mercor.com
APEX-1 + APEX-Agents70.2%2026-10-16reported by mercor.com · mercor.com
Agent Arena+19.5%2026-11-12reported by Latent Space · Latent Space
Agent Duel leaderboard1,0782026-11-19reported by micjm.com · micjm.com
Artificial Analysis Coding Agent Index662026-09-24reported by Latent Space · Latent Space
Artificial Analysis Intelligence Index (adaptive reasoning, max effort, default fallback)582026-09-22reported by artificialanalysis.ai · artificialanalysis.ai
Artificial Analysis Intelligence Index582026-10-23reported by Latent Space, Simon Willison · Simon Willison, Latent Space
Artificial Analysis Intelligence Index592026-11-12reported by Zvi Mowshowitz · Zvi Mowshowitz
Artificial Analysis Intelligence Index v5582026-10-28reported by Latent Space, Zvi Mowshowitz · Latent Space, Zvi Mowshowitz
Artificial Analysis Intelligence Index v5352026-10-31reported by Latent Space · Latent Space
Artificial Analysis Intelligence Index v5592026-11-10reported by Latent Space · Latent Space
BenchCAD 2.165.32026-11-09self-reported · Anthropic the lab's own
CWE-bench v267.3%2026-11-03reported by Hugging Face · Hugging Face
CWE-bench v278.0%2026-11-03reported by Hugging Face · Hugging Face
Chartography64.4%2026-11-09self-reported · Anthropic the lab's own
Coding Agent Index622026-10-29reported by Latent Space · Latent Space
CritPt78%2026-10-23reported by Latent Space · Latent Space
CursorBench 4.052.5%2026-09-22self-reported · Anthropic the lab's own
Cyber ECI1712026-10-23reported by Epoch AI · Epoch AI
DAYJOB: Tax9.2%2026-10-28reported by surgehq.ai · surgehq.ai
DeepSWE74.2%2026-10-01reported by Latent Space · Latent Space
DeepSWE v1.168.4%2026-09-24reported by Latent Space · Latent Space
Epoch Capabilities Index1692026-10-23reported by Epoch AI · Epoch AI
FrontierCode 2.045.5%2026-11-09self-reported · Anthropic the lab's own
FrontierCode v1.1 (Main)54.6%2026-09-22self-reported · Anthropic the lab's own
FrontierCode v1.1 (Main)56.0%2026-10-23reported by Latent Space · Latent Space
GDPx12%2026-11-03reported by Latent Space, Transformer · Latent Space, Transformer
Grey Swan evaluation1 percent2026-10-28reported by Ars Technica · Ars Technica
HLE-Diamond68%2026-10-30reported by Latent Space · Latent Space
HealthBench71.4%2026-10-21reported by Latent Space · Latent Space
HealthBench59.9%2026-10-21reported by Latent Space · Latent Space
Humanity's Last Exam67.7%2026-10-19reported by Import AI · Import AI
Intelligence Index v4.3582026-10-13reported by Latent Space · Latent Space
MASK87.4%2026-09-23reported by Zvi Mowshowitz · Zvi Mowshowitz
MirrorCode77%2026-11-20reported by METR · METR
Models Engineering Index60.82026-11-19reported by git-aarvin.github.io · git-aarvin.github.io
Multimodal Knowledge Agent benchmark53%2026-11-11reported by Latent Space · Latent Space
OSWorld 2.081.8%2026-09-28reported by Hugging Face · Hugging Face
Omniscience462026-09-26reported by Zvi Mowshowitz · Zvi Mowshowitz
ProgramBench18.5%2026-09-26reported by Zvi Mowshowitz · Zvi Mowshowitz
ProgramBench18.5%2026-10-26reported by Import AI · Import AI
ProgramBench70.5%2026-10-26reported by Import AI · Import AI
RSI Index37.32026-11-11reported by Latent Space · Latent Space
SimpleBench88.4%2026-09-25reported by Latent Space · Latent Space
Terminal-Bench 4.063.1%2026-09-24reported by Latent Space · Latent Space
Terminal-Bench-Science62%2026-09-25reported by Latent Space · Latent Space
Terminal-Bench-Science 0.162%2026-09-26reported by Zvi Mowshowitz · Zvi Mowshowitz
ThinkingBox-Bench67.16%2026-10-03reported by Hugging Face · Hugging Face
ThinkingBox-Bench67.162026-10-22reported by Hugging Face · Hugging Face
WeirdML v331.2%2026-09-26reported by Zvi Mowshowitz · Zvi Mowshowitz
political evenhandedness evaluation99%2026-10-14self-reported · Anthropic the lab's own

What it does

In the news

Built 2026-11-21 06:50 UTC from the news knowledge graph by code. Every line is a fact whose quote was checked against the source's own words (hover a source); it says who states it, not that it is true. Archived pages are the lab's pages as the Wayback Machine holds them.