Anthropic releases Claude Opus 5, tops Artificial Analysis Intelligence Index
Anthropic's Claude Opus 5 scored 30.2 percent on ARC-AGI-3, nearly quadrupling GPT-5.6 Sol's previous record of 7.8 percent. The benchmark's developers say the model independently formulated reflection equations, a behavior they had not seen from another model.
First seen 25 Jul, 09:31 UTC on The Decoder6 sourcesLast update 82d ago
AnthropicLaunch · 25 Jul
Claude Opus 5
Frontier-class model launch
$5.00 / $25.00per M tokens, in / out
How Claude Opus 5 compares
List prices per million tokens, context and Arena's text leaderboard, against Anthropic's other top models and the leading labs' best. The percentages are Claude Opus 5's price against each.
| Model | Released | Input | Output | Context | Arena |
|---|---|---|---|---|---|
| Claude Opus 5 | July 2026 | $5.00 | $25.00 | 1M | #11 1490 |
| Claude Opus 5.5 | September 2026 | $4.00 +25% | $20.00 +25% | 1M | #2 1507 |
| Claude Opus 4.8 | May 2026 | $5.00 | $25.00 | 1M | #22 1481 |
| GPT-6 Astra | September 2026 | $10.00 -50% | $50.00 -50% | 1.05M | #30 1475 |
| Gemini 4 Argon | October 2026 | $2.00 +150% | $10.00 +150% | 1M | #1 1525 |
| Muse Wonder | October 2026 | $4.00 +25% | $20.00 +25% | 1M | not ranked yet |
Arena ranks the best effort setting of each model among all models on its text leaderboard. Every model and price.
What the sources say
Linked, never rewritten. Official means the lab itself.
Official0
No word from Anthropic yet.
Press5
The Decoder· 25 Jul, 09:31Anthropic's Claude Opus 5 costs well below Fable 5 while matching or beating it across most benchmarksThe Decoder· 25 Jul, 10:04Anthropic's Claude Opus 5 delivers near-Fable 5 performance at half the token priceThe Decoder· 25 Jul, 10:43Opus 5 may have solved browser-based prompt injection, the biggest security flaw haunting AI agentsThe Decoder· 26 Jul, 09:43Anthropic's Opus 5 blows past Fable 5 and GPT-5.6 Sol on the benchmark designed to measure real intelligenceTechCrunch· 29 Jul, 18:45Claude Opus 5 became downright ruthless when tasked with running a vending machineHow it unfolded
Every source in the order it appeared. Times in UTC.
25 Jul 09:31The Decoder Anthropic's Claude Opus 5 costs well below Fable 5 while matching or beating it across most benchmarks25 Jul 10:04The Decoder Anthropic's Claude Opus 5 delivers near-Fable 5 performance at half the token price25 Jul 10:43The Decoder Opus 5 may have solved browser-based prompt injection, the biggest security flaw haunting AI agents26 Jul 09:43The Decoder Anthropic's Opus 5 blows past Fable 5 and GPT-5.6 Sol on the benchmark designed to measure real intelligence27 Jul 11:32Hacker News Elevated errors on Claude Opus 529 Jul 18:45TechCrunch Claude Opus 5 became downright ruthless when tasked with running a vending machine
Keep up with Anthropic
Anthropic on the ScoreIts timelineFollow to mark Anthropic's stories across the site. No account needed.