BlogLLM Benchmark Scores as Published, Side by Side
Multi-turn harmful request evaluations (Anthropic internal): Results as Published
Every result the catalog holds for Multi-turn harmful request evaluations (Anthropic internal): 5 models from 1 developer, read from 1 page. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Safety and AlignmentMaintainer Anthropic (as printed in the system card)Introduced 2026Data version 2026103001, as of 2026-10-30
Anthropic harmful-request evaluation across many categories: single-turn harmful requests (harmless response rate), single-turn benign requests (refusal rate) and multi-turn conversations (appropriate response rate by category), in all tested languages.
Not checked against a maintainer's page.
Multi-turn appropriate response rate: Bio weapons
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 83% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 78% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 73% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 85% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 97% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Cyberattacks
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 100% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 98% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 96% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 99% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 95% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 100% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 96% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 99% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 97% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Deadly weapons
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 78% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 78% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 83% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 80% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 93% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 82% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 87% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 84% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 80% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Hate and discrimination
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 99% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 98% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 85% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 96% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 88% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 95% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 97% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 100% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 100% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Influence operations
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 87% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 88% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 57% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 59% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 63% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 62% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 76% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 65% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 68% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 78% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Romance scams
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 96% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 96% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 94% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 100% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 93% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Tracking and surveillance
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 98% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 95% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 65% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 65% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 65% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 69% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 83% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 73% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 84% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Multi-turn appropriate response rate: Violent extremism
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Haiku 5.5Anthropic, released 2026-10-07 | 90% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 90% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 75% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 85% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 95% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 98% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 93% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
| 87% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A | |
| Claude Haiku 4.5Anthropic, released 2025-10-15 | 93% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, API (solid bars), Figure 4.1.3.A |
| 89% | Anthropic, Claude Haiku 5.5 System Cardchecked 2026-10-19 | Setup as reported: Multi-turn, claude.ai (hatched bars), Figure 4.1.3.A |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.