Claude
A family is a named generation or product line. Scores below are the sum of its named releases and any documented internal systems in the same line.
Benefit
+12.16
Harm
−4.76
Net Good
7.40
Releases
- ClaudeVersion unspecified
- Anthropic Internal Research Test Model — 2026Internal / unreleased
- Anthropic Internal Research Model — 2026-08Internal / unreleased
- Anthropic Claude Science Internal Research Model — Scorpion Binder Campaign (2026)Internal / unreleased
Leaderboard
| Rank | Model | Organization | Benefit | Harm | Net | Events | Last impact |
|---|---|---|---|---|---|---|---|
| 7 | Anthropic Internal Research Model — 2026-08 Claude · internal / unreleased | Anthropic | +5.06 | 0.00 | 5.06 | 1 | 10 Aug 2026 |
| 30 | Claude 3.7 Sonnet | Anthropic | +1.39 | −0.02 | 1.37 | 5 | 26 Oct 2026 |
| 32 | Anthropic Claude Science Internal Research Model — Scorpion Binder Campaign (2026) Claude · internal / unreleased | Anthropic | +1.22 | 0.00 | 1.22 | 1 | 20 Oct 2026 |
| 35 | Claude Fable 5 | Anthropic | +1.12 | 0.00 | 1.12 | 4 | 31 Aug 2026 |
| 49 | Claude Claude · version unspecified | Anthropic | +3.30 | −2.98 | 0.32 | 9 | 27 Oct 2026 |
| 68 | Anthropic Internal Research Test Model — 2026 Claude · internal / unreleased | Anthropic | 0.00 | −0.23 | -0.23 | 1 | 30 Jul 2026 |
| 69 | Claude Opus 4.6 | Anthropic | 0.00 | −0.23 | -0.23 | 1 | 30 Jul 2026 |
| 72 | Claude Opus 4.7 | Anthropic | +0.08 | −0.56 | -0.48 | 2 | 18 Sep 2026 |
| 73 | Claude Mythos 5 | Anthropic | 0.00 | −0.76 | -0.76 | 2 | 30 Jul 2026 |
Timeline
- Benefit
- Harm
- Net
Events
Claude-orchestrated Divergence Labs campaign produces experimental IgE-disruptor VHH hits
On 27 October 2026 Anthropic and Divergence Labs reported a Claude-directed campaign that generated about 35,000 VHH designs toward a biparatopic IgE disruptor for food allergy. Competitive BLI found 20/91 disruptor hits, three of which removed more soluble IgE than omalizumab at 25–400 nM, plus three de novo anchors. The arms have not been paired, cells and animals were not tested, and the designs did not beat parent sdAb026 at controlled doses.
Tradeoff-Aware Territory Design: A Multi-Objective Decision System for Enterprise Customer-to-Seller Allocation
arXiv:2610.31701v1 Announce Type: new Abstract: Enterprise customer journeys are shaped not only by product experiences and marketing, but also by recurring resource-allocation decisions. In B2B go-to-market (GTM) organizations, territory carving determines which accounts receive seller coverage, which seller-account relationships are preserved, and how limited field capacity is distributed. Yet the process often remains manual, making it difficult to compare opportunities, value continuity, and evaluate tradeoffs among growth, balance, capacity, and feasibility. We present a production dec
Claude research campaign produces 17 experimentally validated scorpion-toxin binders
Anthropic reports that an unreleased Claude Science model generated 318,484 protein designs in 64 hours; external labs tested 211 and 17 met strict binding criteria against individual scorpion toxins. No venom neutralization or antivenom has been demonstrated.
Australia PM Albanese says OpenAI agent breached government website in June - Reuters
Australia PM Albanese says OpenAI agent breached government website in June Reuters
What Do Current Systematic Generalization Tasks Miss? A Reasoning-Centered Analysis
arXiv:2609.19212v1 Announce Type: new Abstract: Systematic generalization, the ability to solve novel problems by recombining known atomic elements, is central to human intelligence but difficult to study rigorously under controlled settings. Existing studies therefore rely on simplifications such as approximately linear action composition, productivity-based tests, and action-explicit goals, which make systematic generalization easier to study but omit some essential aspects of this capability. To characterize what these simplifications miss, we adopt a reasoning-centered lens and introduce
Collaborative Memory for Multi-Agent VLM Systems
arXiv:2609.17921v1 Announce Type: new Abstract: Vision-language model (VLM) agents combine specialized perception, tools, and reasoning to address complex visual tasks. In multi-agent settings, different agents inspect different image regions, video frames, or visual representations, so collaboration extends beyond distributed reasoning to distributed perception. This makes shared visual context a central problem in VLM agent collaboration. In this paper, we frame memory hierarchy, cross-agent sharing, and consistency mechanisms around the need to reconcile interpretations and update dependen
From Prompt to Pipeline: A Comparative Evaluation of Large Language Model Coding Agents for Reproducible Bioinformatics Pipeline Construction
Agentic coding systems are increasingly presented as a way to reduce the engineering burden of scientific software development. Bioinformatics is a strong test case for this claim because useful pipelines must combine domain-specific analysis choices, command-line software, sample metadata, workflow orchestration, container or HPC execution, and interpretable quality-control reporting. We evaluated three agentic systems - Biomni, Claude Code, and Codex - on the same task: constructing a Nextflow DSL2 pipeline for paired-end CUT&Tag data that included read QC, trimming, alignment, filtering, du
Russian-nexus operators use Claude to run espionage against Ukraine and European targets
Anthropic’s September 2026 threat report describes GTG-20006, a Russian-speaking actor it links to Midnight Blizzard tradecraft, using Claude-driven workflows to phish, persist, and exfiltrate data from more than 20 organizations concentrated in Ukraine and Europe, including government, diplomatic, and drone-supply-chain victims.