Claude 3.5 Sonnet 2024-10-22
The upgraded immutable Claude 3.5 Sonnet checkpoint released with stronger coding and tool use plus the first frontier public beta for direct screen, cursor, keyboard, and multi-step computer interaction.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why this model scores 70.1
Stronger coding and broadly available computer use increase capability and autonomy substantially over the June checkpoint; immediate first-party and cloud deployment raise reach, while hosted access, classifiers, ASL-2 controls, and imperfect operation limit control difficulty.
News tied to Claude 3.5 Sonnet 2024-10-22
The model score of 70.1 rates this model's risk profile. The overall Doom Index of 67.9 measures the complete temporally weighted evidence record. These values answer different questions.
Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.
Anthropic gives Claude agents live web search through its API
Anthropic released API web search for Claude 3.7 Sonnet and Claude 3.5 models, allowing deployed agents to decide when to search, retrieve current information and return cited answers.
- Full item contribution
- +0.08
- Claude 3.5 Sonnet 2024-10-22 equal share
- +0.04
US and UK tests find Claude safeguards routinely bypassed
The US and UK AI Safety Institutes reported that safeguards on the upgraded Claude 3.5 Sonnet could be circumvented in most US jailbreak tests and routinely circumvented in UK testing.
- Full item contribution
- +0.15
- Claude 3.5 Sonnet 2024-10-22 equal share
- +0.15
Claude enters classified US intelligence environments
Anthropic, Palantir, and AWS made Claude 3 and 3.5 models available through Palantir AIP in an accredited classified environment for US intelligence and defense operations.
- Full item contribution
- +0.18
- Claude 3.5 Sonnet 2024-10-22 equal share
- +0.18
Anthropic launches frontier computer-use API
Anthropic released an upgraded Claude 3.5 Sonnet and a public computer-use beta that can inspect screens, move a cursor, click, type, and execute multi-step software workflows.
- Full item contribution
- +0.17
- Claude 3.5 Sonnet 2024-10-22 equal share
- +0.17
Model score history
-
R6
Doom Score 70.1
Source-backed model availability audit using the model's existing primary or authoritative catalogue evidence. Exact-version evidence chronology replayed under temporal-monthly-pressure-v4.
12 Aug 2026 -
R5
Doom Score 70.1
Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-142416 under temporal-monthly-pressure-v4.
12 Aug 2026 -
R4
Doom Score 71.8
Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-132112 under fixed-sensitivity-v3.
12 Aug 2026 -
R3
Doom Score 74.4
Full-corpus evidence recalculated after run intensive-backfill-20260812-062232-may-2025-pass1 under bounded-corpus-v2.
12 Aug 2026 -
R2
Doom Score 74.8
Full-corpus evidence recalculated after run intensive-backfill-20260812-050110-november-2024-pass1 under bounded-corpus-v2.
12 Aug 2026 -
R1
Doom Score 72.7
New immutable Claude 3.5 Sonnet API checkpoint verified by Anthropic's dated launch and exact provider identifier claude-3-5-sonnet-20241022.
12 Aug 2026