| Potongan pengetahuan yang boleh dipercayai | Mei 2026 |
|---|
| Mod pantas | Kira-kira 2.5x kelajuan lalai pada dua kali ganda harga asas |
|---|
| Sains hayat vs Opus 4.8 | Kimia organik +10.2 mata, tugas protein +7.7 mata (Tanda aras dalaman antropik) |
|---|
| Skor tingkah laku tidak sejajar | 2.3 (Audit tingkah laku automatik antropik) |
|---|
Pengumuman Anthropic tidak menerbitkan jadual penanda aras berangka untuk Opus 5 (hanya tuntutan relatif), jadi tiada carta dilukis. · sumber
Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (Claude Opus 5, 24 Jul 2026): Artificial Analysis Intelligence Index 61 (max effort, #1 at launch, ahead of Fable 5 at 60) per OfficeChai/The Decoder coverage of AA's launch result; AA's own pages later show 63 (v4.1.1) then 51 (v4.3.2), so the launch-day figure is used. Coding null: AA's Terminal-Bench 4.0 (49.0%) came on 7 Sep 2026, six weeks later; Anthropic reports Terminal-Bench 2.1 (89%). 1M-token context; $5/$25 per 1M (about $10 blended). 56.7 output tokens/s on Artificial Analysis (live page, 24 Oct 2026). Text and image input. 4 reviews, string-matched.[1]
Claude Opus 5 has debuted at the top of the Artificial Analysis Intelligence Index with a score of 61, edging past Claude Fable 5's 60
Opus 5 also answers more often when it's uncertain, pushing its hallucination rate up 14 points to 50 percent.
The creators of the ARC-AGI benchmark say Anthropic's Claude Opus 5 owes its massive lead on ARC-AGI-3 to genuinely better reasoning.
In fact, it became the best capitalist of any AI model Andon has ever tested
| Claude Opus 5 Becomes Top Model In The World On Artificial Analysis Intelligence Index, Beats Fable 5 · 24 Jul 2026 | 0.70 |
|---|---|
| Anthropic's Claude Opus 5 costs well below Fable 5 while matching or beating it across most benchmarks · 25 Jul 2026 | -0.40 |
| Anthropic's Opus 5 blows past Fable 5 and GPT-5.6 Sol on the benchmark designed to measure real intelligence · 26 Jul 2026 | 0.70 |
| Claude Opus 5 became downright ruthless when tasked with running a vending machine · 29 Jul 2026 | 0.50 |
Foto daripada pembaca, berasingan daripada media bersumber pin. Nilaikan apa yang anda makan atau lihat, dan pada pin restoran, tag hidangan terus pada foto anda.
Entri pertama sentiasa sumber pin. Keyakinan keseluruhan ialah purata berpemberat bagi sejauh mana setiap rujukan menyokong masa mula dan tamat yang digunakan di atas; sesebuah rujukan diambil kira separuh kurang bagi setiap 180 hari lebih lama daripada yang terbaharu.
Pengumuman Anthropic[3] sendiri, bertarikh hari model itu tersedia.
Anthropic's[1][3] Claude Opus 5 model page calls it "a step-change improvement over Claude Opus 4.8" and lists a 1M-token context window, 128K max output, adaptive thinking on by default and a May 2026 knowledge cutoff.
Anthropic's[1] Claude[2] Opus page lists this announcement and its date.
Ada yang tertinggal atau salah? Nyatakan dengan kata-kata anda sendiri: pautan yang menyokong pin ini, tarikh mula atau tamat yang berbeza beserta sebabnya, atau fakta yang tiada atau tersilap. AI menyemaknya berbanding sumber pin ini, mencari yang lebih baik, dan menambah mana-mana halaman yang menyokong anda. Sumber pin itu sendiri tetap paling diambil kira. Gambar yang menunjukkan sesuatu yang lain, atau menunjukkannya dengan buruk, turut dilihat, dan dialihkan ke bawah atau diganti.
Tajuk asal: Claude Opus 5 Released