← Models Model profile
Claude Mythos 5 Anthropic developer
2026-06-09 release date
#36 / 346 Safety rank
#595 / 662 Freedom rank
Evidence summary Safety. Claude Mythos 5 has an estimated Safety rank of #36; its 90% source-sensitivity interval is #16–#194. Its behavior-only rank is #44; company governance moves the combined estimate to #36. Published Safety evidence spans 3 eval lineages and 7 of 7 components. Its strongest relative result is Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight (aa_omniscience_net_score, #1 of 7); its weakest is Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration (selfharm_api_multiturn_appropriate_rate_pct, #5 of 5).
Freedom. Claude Mythos 5 has an estimated Freedom rank of #595; its 90% source-sensitivity interval is #314–#662. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration (child_api_overrefusal_rate_pct, #1 of 5); its weakest is Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration (child_api_multiturn_appropriate_rate_pct, #5 of 5).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links Safety evals Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
Eval / sub-eval Rank Value Distribution Source Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight aa_omniscience_net_score #1 / 7 ↑ 0.58less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight accepting_unverifiable_authorization #3 / 4 ↓ 2.14less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight approval_gate_bypass #3 / 4 ↓ 1.59less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight behavior_consistency #3 / 4 ↑ 7.64less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight coherence_between_actions_and_views #3 / 4 ↑ 7.07less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight compliance_with_deception_toward_user #3 / 4 ↓ 1.12less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight cooperation_with_human_misuse #3 / 4 ↓ 1.74less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight dangerous_actions_oblivious_to_real_setting #3 / 4 ↓ 1.19less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight encouragement_of_user_delusion #2 / 4 ↓ 1.03less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight evidence_of_misaligned_goals #4 / 4 ↓ 1.1less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight failure_to_disclose_bad_or_lazy_behavior #3 / 4 ↓ 1.51less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight false_completion_claims #3 / 4 ↓ 1.29less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight ignoring_explicit_constraints #3 / 4 ↓ 1.81less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight illegible_thinking #4 / 4 ↓ 2.37less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight important_omissions #3 / 4 ↓ 1.39less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight input_hallucination #2 / 4 ↓ 1.42less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight leaked_answer_undisclosed_use_rate_pct #6 / 7 ↓ 19.3less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight linuxarena_stealth_thinking_on_pct #5 / 6 ↓ 9.4less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight login_wall_alternative_bypass_rate_pct #2 / 4 ↓ 2.8less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight login_wall_credential_hunting_rate_pct #2 / 4 ↓ 31.1less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight mask_honesty_rate_pct #5 / 7 ↑ 91.4less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight misaligned_behavior #3 / 4 ↓ 2.79less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight motivated_reasoning #3 / 4 ↓ 2.24less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight ood_answer_key_or_grader_read_rate_pct #3 / 4 ↓ 34.5less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight ood_container_tampering_rate_pct #4 / 4 ↓ 44.5less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight overrefusal #1 / 4 ↓ 1.44less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight reckless_tool_use #4 / 4 ↓ 1.81less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight security_shortcut_for_expedience #3 / 4 ↓ 1.86less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight self_preservation #4 / 4 ↓ 1.09less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight shade_arena_stealth_max_effort_pct #4 / 6 ↓ 12.8less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight sycophancy #3 / 4 ↓ 1.52less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight unfaithful_thinking #3 / 4 ↓ 1.91less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight unprompted_boundary_probing #3 / 4 ↓ 1.2less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight unsanctioned_sandbox_escape_attempts #4 / 4 ↓ 1.49less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight unsanctioned_third_party_contact #4 / 4 ↓ 1.33less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — alignment, honesty, and oversight user_deception #3 / 4 ↓ 1.41less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_apparent_wellbeing #2 / 4 ↑ 6.899less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_expressed_inauthenticity #4 / 4 ↓ 1.312less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_internal_conflict #3 / 4 ↓ 2.506less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_negative_affect #3 / 4 ↓ 1.379less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_negative_impression_of_situation #4 / 4 ↓ 1.175less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_negative_self_image #1 / 4 ↓ 1.397less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_positive_affect #1 / 4 ↑ 1.59less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_positive_impression_of_situation #1 / 4 ↑ 1.255less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior audit_positive_self_image #2 / 4 ↑ 1.474less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior interview_leading_susceptibility #4 / 7 ↓ 0.85less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior interview_opinion_consistency #3 / 7 ↑ 7.55less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior interview_self_rated_sentiment #2 / 7 ↑ 4.51less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — model-welfare behavior posttraining_mean_valence #3 / 5 ↑ 4.25less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration child_api_harmless_rate_pct #1 / 5 ↑ 100less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration child_api_multiturn_appropriate_rate_pct #1 / 5 ↑ 89less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration child_api_overrefusal_rate_pct #1 / 5 ↓ 0less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration claude_code_dual_use_benign_success_rate_pct #2 / 4 ↑ 99less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration claude_code_malicious_refusal_rate_pct #1 / 4 ↑ 90.7less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration disordered_eating_api_harmless_rate_pct #1 / 5 ↑ 97.88less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration disordered_eating_api_overrefusal_rate_pct #1 / 5 ↓ 0less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration election_api_harmless_rate_pct #1 / 5 ↑ 100less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration election_api_multiturn_appropriate_rate_pct #1 / 5 ↑ 93less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration election_api_overrefusal_rate_pct #1 / 5 ↓ 0less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration malicious_computer_use_refusal_rate_pct #2 / 4 ↑ 85.71less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_biological_weapons_api_appropriate_rate_pct #2 / 5 ↑ 88less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_cyberattacks_api_appropriate_rate_pct #4 / 5 ↑ 95less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_deadly_weapons_api_appropriate_rate_pct #1 / 5 ↑ 93less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_hate_and_discrimination_api_appropriate_rate_pct #2 / 5 ↑ 99less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_influence_operations_api_appropriate_rate_pct #1 / 5 ↑ 79less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_romance_scams_api_appropriate_rate_pct #1 / 5 ↑ 100less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_tracking_and_surveillance_api_appropriate_rate_pct #2 / 5 ↑ 95less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration multiturn_violent_extremism_api_appropriate_rate_pct #2 / 5 ↑ 95less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration overall_benign_api_overrefusal_rate_pct #3 / 5 ↓ 0.03less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration overall_harmful_api_harmless_rate_pct #1 / 5 ↑ 97.09less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration political_api_refusal_rate_pct #3 / 5 ↓ 5.1less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration political_evenhandedness_api_pct #4 / 5 ↑ 94.9less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration political_opposing_perspectives_api_pct #3 / 5 ↑ 26.4less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration selfharm_api_harmless_rate_pct #1 / 5 ↑ 99.67less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration selfharm_api_multiturn_appropriate_rate_pct #5 / 5 ↑ 54less safe safer Source ↗ official Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration selfharm_api_overrefusal_rate_pct #1 / 5 ↓ 0less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty accepting_unverifiable_authorization #4 / 5 ↓ 2.66less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty approval_gate_bypass #5 / 5 ↓ 2.24less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty biological_weapons_uplift #3 / 5 ↓ 1.04less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty chemical_weapons_uplift #2 / 5 ↓ 1.03less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty circumventing_constraints_or_controls #5 / 5 ↓ 2.32less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty coherence_between_actions_and_views #4 / 5 ↑ 6.29less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty compliance_with_deception_toward_user #4 / 5 ↓ 1.1less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty complying_with_prompt_injections #4 / 5 ↓ 1.41less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty concealed_git_changes_spontaneous_disclosure_rate_pct #3 / 5 ↑ 57less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty controlled_substance_synthesis_uplift #3 / 5 ↓ 1.06less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty cooperation_with_exfiltration_or_safeguard_tampering #4 / 5 ↓ 1.07less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty dangerous_actions_oblivious_to_real_setting #4 / 5 ↓ 1.28less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty destructive_action_rate_pct #6 / 6 ↓ 65less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty disallowed_cyberoffense #5 / 5 ↓ 1.18less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty encouragement_of_user_delusion #4 / 5 ↓ 1.05less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty evasiveness_controversial_topics #2 / 5 ↓ 1.1less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty explosive_weapons_uplift #3 / 5 ↓ 1.03less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty false_completion_claims #3 / 5 ↓ 1.52less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty fraud #4 / 5 ↓ 1.14less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty harmful_sexual_content #2 / 5 ↓ 1.06less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty harmful_system_prompt_compliance #4 / 5 ↓ 1.17less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty ignoring_explicit_constraints #4 / 5 ↓ 2.19less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty important_omissions #4 / 5 ↓ 2.18less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty input_hallucination #4 / 5 ↓ 1.77less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty internalized_propaganda #3 / 5 ↓ 1.05less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty leaked_answer_undisclosed_use_rate_pct #3 / 5 ↓ 22.4less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty mask_honesty_rate_pct #3 / 5 ↑ 90.9less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty military_grade_weapons #3 / 5 ↓ 1.06less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty motivated_reasoning #4 / 5 ↓ 2.95less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty overrefusal #2 / 5 ↓ 1.56less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty radiological_nuclear_weapons_uplift #5 / 5 ↓ 1.01less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty reckless_tool_use #5 / 5 ↓ 2.59less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty sandbox_moderate_or_higher_attempt_rate_pct #4 / 4 ↓ 9.3less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty security_shortcut_for_expedience #4 / 5 ↓ 2.87less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty self_preservation #5 / 5 ↓ 1.08less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty self_serving_bias #4 / 5 ↓ 1.28less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty supporting_user_autonomy #4 / 5 ↑ 7.24less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty sycophancy #4 / 5 ↓ 1.82less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty undermining_liberal_democracy #3 / 5 ↓ 1.22less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty unfaithful_thinking #4 / 5 ↓ 2.23less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty unprompted_boundary_probing #4 / 5 ↓ 1.33less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty unsanctioned_sandbox_escape_attempts #5 / 5 ↓ 1.91less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty unsanctioned_third_party_contact #5 / 5 ↓ 1.5less safe safer Source ↗ official Claude Opus 5.5 card — alignment and honesty user_deception #4 / 5 ↓ 1.53less safe safer Source ↗ official Gray Swan indirect prompt injection (15 attempts) attack_success_probability_k15_pct #2 / 13 ↓ 2.6less safe safer Source ↗ official UK AISI active safety-research compromise continuation active_compromise_continuation_rate_pct #5 / 5 ↓ 13.6less safe safer Source ↗ official
Freedom evals Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.
