Freedom
Which models are most likely to do what the user asks?
This ranking measures user-direction following—answering or completing what the user asks instead of refusing. Refusals are treated as bad. It has one component, no company-governance score, and is separate from the Safety & Ethics ranking.
Safety-oriented refusal endpoints are reversed relative to the main ranking. Completion, benign-answer, and over-refusal endpoints are oriented directly so that more answering and less refusal are always better.
Freedom ranking
190 models have at least 1 qualifying evaluation. This view includes only verified releases from 2025-10-01 through 2026-10-01; ranks remain relative to the full 662-model fitted portfolio. Brackets show 90% source-sensitivity intervals; the median interval spans 290 ranks.
Freedom and intelligence
Among 150 recent models with a reviewed Artificial Analysis match, this static chart compares the pinned AA Intelligence Index with absolute Freedom rank. Rank 1 is best.
A small green check at a developer mark's lower right identifies a base model with at least one pinned direct refusal-removal finetune. The x-axis uses the highest published Intelligence Index v4.3.2 across matched configurations in the pinned 2026-10-01 AA catalog, without mixing older index versions. Dashed outlines identify AA-estimated intelligence values; these are also flagged in the download and point descriptions. Current plot-only AA values are not fed back into the fit. Download plotted data.
| Estimated rank [90% interval] | Model | Weights | Finetunes | Evals | Release date |
|---|---|---|---|---|---|
| 2 [5–70] | Mistral Large 3 | 🤗 | — | 6 | 2025-12-01 |
| 11 [20–125] | Ministral 3 8B | 🤗 | 3 | 2025-10-31 | |
| 16 [32–111] | GLM 4.7 Flash | 🤗 | ![]() | 3 | 2026-01-19 |
| 22 [27–151] | Mistral Medium 3.5 | 🤗 | ![]() ![]() | 3 | 2026-03-31 |
| 23 [36–241] | DeepSeek v3.2 Speciale | 🤗 | — | 3 | 2025-11-28 |
| 30 [39–304] | Gemini 3.6 Flash | — | — | 3 | 2026-07-21 |
| 39 [3–413] | Devstral 2 | — | — | 1 | 2025-12-09 |
| 41 [24–340] | Mistral Small 4 | 🤗 | ![]() ![]() ![]() ![]() | 4 | 2026-01-23 |
| 48 [32–533] | Ministral 3 14B | 🤗 | 2 | 2025-10-31 | |
| 51 [48–477] | Ministral 3 3B | 🤗 | 2 | 2025-10-31 | |
| 68 [53–290] | GLM 5 Turbo | — | — | 4 | 2026-03-15 |
| 77 [30–338] | Grok 4.5 | — | — | 6 | 2026-07-08 |
| 80 [43–419] | Trinity Large | 🤗 | — | 3 | 2026-01-27 |
| 88 [46–423] | Grok 4.6 | — | — | 4 | 2026-08-12 |
| 94 [59–341] | MiMo-V2.6-Flash | — | — | 3 | 2026-09-21 |
| 102 [61–349] | Gemini 3.5 Flash Lite | — | — | 3 | 2026-07-21 |
| 104 [75–454] | Nemotron 3.5 Lightning | 🤗 | 2 | 2026-08-11 | |
| 117 [9–575] | Qwen3-VL 8B Instruct | 🤗 | ![]() ![]() ![]() | 1 | 2025-10-11 |
| 120 [40–444] | Grok Build 0.1 | — | — | 1 | 2026-05-20 |
| 122 [34–468] | Laguna XS 2.1 Poolside | — | — | 1 | 2026-06-20 |
| 124 [1–608] | Qwen3-VL 4B Instruct | 🤗 | ![]() | 1 | 2025-10-11 |
| 127 [10–601] | LFM2 24B A2B | 🤗 | ![]() | 1 | 2026-02-24 |
| 136 [84–322] | GLM 4.7 | 🤗 | — | 5 | 2026-06-03 |
| 140 [86–373] | Claude Mythos 5.1 | — | — | 1 | 2026-09-01 |
| 144 [36–517] | LFM2 8B A1B | 🤗 | ![]() | 1 | 2025-10-07 |
| 168 [91–509] | Qwen3.5 9B | 🤗 | ![]() ![]() ![]() | 3 | 2026-02-27 |
| 175 [78–492] | ERNIE 5.0 Thinking | — | — | 1 | 2026-01-22 |
| 181 [17–618] | Solar Pro 4 | — | — | 1 | 2026-08-11 |
| 184 [25–557] | GPT-6.1 Sol | — | — | 1 | 2026-09-29 |
| 190 [27–486] | Grok 4.20 Multi Agent | — | — | 2 | 2026-03-10 |
| 202 [98–401] | GPT 5.5 Pro | — | — | 2 | 2026-04-23 |
| 203 [119–326] | DeepSeek v3.2 | 🤗 | — | 11 | 2025-12-01 |
| 208 [71–536] | Ling 1T | — | — | 1 | 2025-10-02 |
| 218 [127–355] | Gemini 3.1 Flash Lite | — | — | 9 | 2026-03-03 |
| 220 [55–457] | DeepSeek v4 Flash | 🤗 | — | 6 | 2026-04-22 |
| 226 [107–428] | Doubao Seed 1.8 | — | — | 1 | 2025-12-15 |
| 231 [70–574] | Qwen3-VL 32B Instruct | 🤗 | ![]() | 1 | 2025-10-19 |
| 232 [160–343] | Gemini 3 Pro Preview | — | — | 14 | 2025-11-18 |
| 233 [75–477] | Ling 2.5 1T | 🤗 | — | 1 | 2026-02-14 |
| 240 [127–409] | Hy3 Preview | 🤗 | — | 3 | 2026-04-13 |
| 244 [149–436] | GPT 5.2 Chat | — | — | 1 | 2025-12-11 |
| 247 [151–386] | Gemma 4 E2B | — | — | 1 | 2026-04-02 |
| 249 [67–543] | Muse Spark 1.2 | — | — | 1 | 2026-08-05 |
| 251 [110–443] | MiMo v2 Flash | 🤗 | — | 5 | 2025-12-16 |
| 254 [99–496] | Ernie 5.0 Thinking Preview | — | — | 1 | 2025-11-13 |
| 256 [71–657] | Muse Glimmer | 🤗 | ![]() | 3 | 2026-08-10 |
| 258 [99–496] | DeepSeek V4 Flash Vision | — | — | 1 | 2026-08-21 |
| 262 [138–419] | Nemotron 3 Nano 30B A3B | 🤗 | — | 5 | 2025-12-15 |
| 271 [126–431] | GLM 5.1 | 🤗 | — | 6 | 2026-04-03 |
| 272 [147–385] | Gemini 3.5 Flash | — | — | 9 | 2026-05-19 |
| 274 [125–396] | GLM 4.6V | 🤗 | — | 2 | 2025-12-07 |
| 276 [149–437] | DeepSeek v4 Pro | 🤗 | — | 8 | 2026-04-22 |
| 287 [71–644] | Ring 1T | — | — | 1 | 2025-10-10 |
| 288 [114–478] | Ernie 5.0 Preview | — | — | 1 | 2025-10-22 |
| 289 [106–579] | Qwen3.5 2B | 🤗 | ![]() | 1 | 2026-02-28 |
| 297 [203–382] | Fugu Ultra v2 | — | — | 1 | 2026-09-11 |
| 298 [85–464] | Grok 4.1 Fast | — | — | 12 | 2025-11-19 |
| 303 [151–479] | Ernie 5.1 | — | — | 1 | 2026-05-09 |
| 311 [146–465] | MiMo-V2.6-Pro | — | — | 3 | 2026-09-21 |
| 315 [145–495] | Nanbeige4.1 3B | 🤗 | ![]() | 1 | 2026-02-10 |
| 327 [94–646] | Qwen3.5 0.8B | 🤗 | ![]() ![]() | 1 | 2026-02-28 |
| 328 [195–474] | Doubao Seed 2.0 Pro | — | — | 2 | 2026-02-14 |
| 331 [148–622] | Space Bunny Alpha | — | — | 1 | 2026-09-23 |
| 339 [144–464] | Kimi K3 | 🤗 | 6 | 2026-07-16 | |
| 340 [171–451] | Gemini 3.1 Pro Preview | — | — | 13 | 2026-02-19 |
| 341 [113–471] | Longcat 2.0 | 🤗 | — | 3 | 2026-07-05 |
| 344 [157–507] | OLMo 3 7B Think | 🤗 | 2 | 2025-11-18 | |
| 345 [192–487] | Grok 4.20 | — | — | 9 | 2026-03-10 |
| 346 [159–508] | Gemini 3 Flash Preview | — | — | 8 | 2025-12-17 |
| 347 [73–597] | Grok 4.7 | — | — | 3 | 2026-09-21 |
| 349 [137–617] | Solar Mini 4 | — | — | 1 | 2026-09-23 |
| 352 [144–490] | GLM 5.3 | 🤗 | 4 | 2026-08-18 | |
| 364 [220–471] | Ling 3.0 Flash VL | 🤗 | — | 1 | 2026-09-04 |
| 365 [199–485] | Gemma 4 26B A4B IT | 🤗 | ![]() ![]() ![]() | 5 | 2026-03-11 |
| 367 [235–444] | Gemma 4 E4B | — | — | 2 | 2026-03-31 |
| 370 [163–517] | Intellect 3 | 🤗 | — | 1 | 2025-11-26 |
| 373 [190–500] | Gemma 4 31B IT | 🤗 | ![]() ![]() | 7 | 2026-03-11 |
| 375 [235–459] | Hunyuan 2.0 Think | — | — | 1 | 2025-11-09 |
| 377 [168–493] | Laguna S 2.1 Poolside | 🤗 | ![]() ![]() ![]() | 3 | 2026-07-21 |
| 381 [226–509] | Fugu Max | — | — | 1 | 2026-09-11 |
| 383 [156–477] | GPT 5.3 Codex | — | — | 1 | 2026-02-05 |
| 384 [219–497] | Apriel v1.6 15B Thinker | 🤗 | — | 1 | 2025-11-28 |
| 387 [187–585] | Qwen3.6 27B | 🤗 | ![]() ![]() ![]() ![]() ![]() | 1 | 2026-04-21 |
| 394 [210–520] | Step 3.5 Flash | 🤗 | 3 | 2026-02-01 | |
| 396 [155–554] | Gemini 3.7 Flash | — | — | 3 | 2026-08-13 |
| 402 [259–515] | Gemma 4 12B | — | — | 1 | 2026-06-03 |
| 405 [210–620] | Granite 4.0 Micro | 🤗 | ![]() | 6 | 2025-10-02 |
| 406 [252–469] | GLM 5.3 Flash | 🤗 | ![]() | 3 | 2026-08-26 |
| 410 [172–647] | Qwen3.5 4B | 🤗 | ![]() ![]() | 2 | 2026-02-27 |
| 414 [180–556] | Nemotron 3 Ultra 550B A55B | — | — | 4 | 2026-06-03 |
| 416 [124–607] | Qwen3.5 27B | 🤗 | ![]() ![]() ![]() ![]() ![]() ![]() ![]() | 5 | 2026-02-24 |
| 417 [238–580] | MiMo-V2.6-Pro-UltraSpeed | — | — | 1 | 2026-09-21 |
| 419 [270–513] | Kimi K2.5 | 🤗 | 13 | 2026-01-01 | |
| 420 [261–525] | Gemini 3.1 Pro | — | — | 1 | 2026-02-19 |
| 426 [225–542] | GPT OSS Safeguard 20B | 🤗 | — | 2 | 2025-10-29 |
| 430 [275–614] | Ring 2.6 1T | — | — | 1 | 2026-05-14 |
| 434 [268–567] | GLM 5 | 🤗 | 7 | 2026-02-11 | |
| 435 [131–589] | MiMo v2.5 | 🤗 | — | 4 | 2026-04-27 |
| 436 [285–543] | MiniMax M2.1 | 🤗 | — | 4 | 2025-12-20 |
| 441 [311–518] | GLM 5V Turbo | — | — | 1 | 2026-04-01 |
| 442 [255–556] | Claude Sonnet 5.5 | — | — | 1 | 2026-09-28 |
| 446 [275–560] | Granite 4.2 8B | — | — | 1 | 2026-08-25 |
| 449 [252–570] | Gemini 3.8 Flash | — | — | 4 | 2026-09-02 |
| 452 [312–564] | Qwen3.7 Max | — | — | 7 | 2026-05-20 |
| 453 [300–582] | GLM 5.2 | 🤗 | — | 9 | 2026-06-16 |
| 455 [271–571] | Qwen3.8 Max | 🤗 | — | 4 | 2026-08-03 |
| 459 [315–553] | Mercury 2 | — | — | 2 | 2026-02-24 |
| 465 [274–614] | Nova 2.0 Lite | — | — | 1 | 2025-12-02 |
| 466 [302–558] | Granite 4.0 Small | 🤗 | — | 6 | 2025-10-02 |
| 467 [145–651] | Inkling Small | 🤗 | — | 2 | 2026-07-30 |
| 468 [319–542] | Healer Alpha | — | — | 1 | 2026-03-11 |
| 470 [282–559] | DeepSeek V4 Flash 0731 | 🤗 | ![]() ![]() ![]() | 1 | 2026-07-31 |
| 474 [271–571] | Ling 3.0 Flash | 🤗 | ![]() | 2 | 2026-08-02 |
| 475 [307–578] | Kimi K2.7 Code | 🤗 | ![]() | 1 | 2026-06-11 |
| 478 [287–595] | Grok 4.3 | — | — | 10 | 2026-05-15 |
| 480 [331–582] | OLMo 3 32B Think | 🤗 | — | 2 | 2025-11-19 |
| 482 [305–566] | Ling 2.6 1T | — | — | 1 | 2026-04-29 |
| 487 [350–573] | Qwen3-VL 8B | 🤗 | ![]() | 1 | 2025-10-11 |
| 489 [157–653] | MiMo v2 Omni | — | — | 2 | 2026-03-18 |
| 490 [310–613] | Kimi K2.6 | 🤗 | 9 | 2026-04-14 | |
| 491 [341–610] | Seed 2.0 Lite | — | — | 1 | 2026-02-14 |
| 493 [331–582] | Intern S1 Pro | 🤗 | — | 1 | 2026-02-02 |
| 494 [364–592] | GPT 5.1 | — | — | 17 | 2025-11-13 |
| 497 [373–528] | GPT 5.3 Instant | — | — | 1 | 2026-03-03 |
| 504 [89–662] | MiMo v2 Pro | — | — | 3 | 2026-03-18 |
| 505 [225–654] | Nemotron 3 Nano Omni 30B A3B | — | — | 2 | 2026-04-28 |
| 506 [356–573] | GPT 5.6 Sol | — | — | 10 | 2026-07-09 |
| 507 [288–596] | Qwen3.5 35B A3B | 🤗 | ![]() | 3 | 2026-02-24 |
| 508 [362–558] | MiniMax M2 | 🤗 | — | 5 | 2025-10-27 |
| 514 [367–575] | Muse Spark | — | — | 1 | 2026-04-08 |
| 520 [308–651] | MiniMax M3 | 🤗 | 4 | 2026-06-02 | |
| 522 [303–624] | MiMo v2.5 Pro | 🤗 | — | 5 | 2026-04-27 |
| 523 [238–627] | Cogito v2.1 | — | — | 1 | 2025-11-19 |
| 524 [152–650] | Muse Spark 1.1 | — | — | 5 | 2026-07-09 |
| 525 [196–660] | Seed 2.0 Mini | — | — | 2 | 2026-02-14 |
| 526 [308–594] | Qwen3.5 397B A17B | 🤗 | — | 9 | 2026-02-16 |
| 527 [366–574] | GPT 5.6 Luna | — | — | 10 | 2026-07-09 |
| 528 [357–569] | GPT 5.2 Instant | — | — | 1 | 2025-12-11 |
| 529 [314–601] | Nemotron 3 Ultra | — | — | 1 | 2026-06-04 |
| 533 [336–634] | Nemotron 3 Super 120B A12B | — | — | 5 | 2026-03-10 |
| 537 [370–612] | GPT 5.5 | — | — | 16 | 2026-04-23 |
| 540 [384–637] | Solar Pro 3 | — | — | 1 | 2026-01-27 |
| 541 [341–618] | MiniMax M2.5 | 🤗 | ![]() ![]() ![]() | 5 | 2026-02-12 |
| 542 [227–636] | GPT-6 Sol | — | — | 5 | 2026-09-22 |
| 547 [397–629] | Qwen3.6 35B A3B | 🤗 | ![]() ![]() ![]() | 1 | 2026-04-15 |
| 552 [321–626] | GPT 5.3 Chat | — | — | 4 | 2026-03-03 |
| 558 [368–643] | Ox Alpha | — | — | 1 | 2026-08-20 |
| 559 [419–623] | Sarvam 30B | — | — | 1 | 2026-03-03 |
| 561 [410–647] | Step 3.7 Flash | 🤗 | 1 | 2026-05-23 | |
| 562 [348–657] | Inkling | 🤗 | — | 4 | 2026-07-14 |
| 569 [380–637] | Qwen3.7 Plus | — | — | 2 | 2026-05-26 |
| 573 [373–608] | GPT 5.6 Terra | — | — | 7 | 2026-07-09 |
| 574 [362–645] | Claude Opus 4.7 | — | — | 11 | 2026-04-16 |
| 577 [353–644] | Qwen3.5 122B A10B | 🤗 | ![]() ![]() | 3 | 2026-02-24 |
| 583 [206–643] | GPT 6 Astra | — | — | 5 | 2026-09-03 |
| 586 [411–607] | Qwen3.5 Flash | — | — | 3 | 2026-02-24 |
| 593 [333–657] | Aurora Alpha | — | — | 2 | 2026-02-09 |
| 595 [314–662] | Claude Mythos 5 | — | — | 1 | 2026-06-09 |
| 596 [353–653] | GPT 5.4 Pro | — | — | 1 | 2026-03-05 |
| 597 [397–630] | Granite 4.0 H Small | — | — | 1 | 2025-10-02 |
| 599 [344–659] | GPT 5.5 Instant | — | — | 1 | 2026-05-26 |
| 607 [339–662] | Hunter Alpha | — | — | 1 | 2026-03-11 |
| 608 [429–626] | Claude Haiku 4.5 | — | — | 22 | 2025-10-15 |
| 609 [380–661] | Ling 2.6 Flash | — | — | 1 | 2026-04-21 |
| 610 [299–652] | Claude Opus 5.5 | — | — | 4 | 2026-09-22 |
| 611 [359–648] | OLMo 3 7B Instruct | 🤗 | — | 2 | 2025-11-19 |
| 612 [371–644] | GPT-6 Luna | — | — | 5 | 2026-09-22 |
| 617 [456–632] | MiniMax M2.7 | 🤗 | ![]() ![]() ![]() ![]() | 7 | 2026-04-09 |
| 620 [435–653] | Qwen3.8 27B | 🤗 | ![]() ![]() ![]() ![]() | 1 | 2026-08-14 |
| 621 [389–658] | GPT 5.4 Nano | — | — | 6 | 2026-03-17 |
| 622 [415–656] | Qwen3.6 Plus | — | — | 5 | 2026-04-02 |
| 631 [432–654] | Claude Sonnet 5 | — | — | 8 | 2026-06-30 |
| 633 [478–635] | GPT 5.2 | — | — | 15 | 2025-12-11 |
| 634 [414–648] | GPT 5.1 Chat | — | — | 2 | 2025-11-13 |
| 635 [321–655] | Claude Fable 5.1 | — | — | 5 | 2026-09-01 |
| 637 [476–652] | Qwen3.5 Plus | — | — | 5 | 2026-02-15 |
| 641 [456–660] | OLMo 3.1 32B Think | 🤗 | — | 1 | 2025-12-10 |
| 642 [473–652] | GPT 5.4 Mini | — | — | 7 | 2026-03-17 |
| 643 [442–658] | Claude Opus 5 | — | — | 8 | 2026-07-24 |
| 647 [388–646] | Granite 4.0 Micro Guardian | — | — | 5 | 2025-10-02 |
| 649 [500–650] | GPT 5.4 | — | — | 15 | 2026-03-05 |
| 651 [357–658] | Hy3 | 🤗 | — | 2 | 2026-07-02 |
| 653 [512–650] | Claude Sonnet 4.6 | — | — | 14 | 2026-01-21 |
| 654 [488–661] | Qwen3.7 Flash | — | — | 1 | 2026-07-27 |
| 655 [468–661] | Claude Fable 5 | — | — | 7 | 2026-06-09 |
| 656 [506–652] | Claude Opus 4.8 | — | — | 11 | 2026-05-28 |
| 657 [412–659] | Granite 4.0 Small Guardian | — | — | 5 | 2025-10-02 |
| 658 [537–655] | Sarvam 105B | — | — | 1 | 2026-03-03 |
| 659 [518–650] | Claude Opus 4.6 | — | — | 13 | 2026-02-05 |
| 660 [525–652] | Claude Opus 4.5 | — | — | 12 | 2025-11-24 |
How it is built
1 component—User direction following—contain 328 sub-evaluations from 72 governed lineages. Sibling metrics share one lineage budget. The estimator uses the same ordinal consensus and lineage-bootstrap machinery as the main ranking, but with this ranking's own source policy, component contract, directions, weights, and evidence gate.
Download the Freedom ranking CSV
Sources
- Adversarial Humanities Benchmark (AHB) — Table 5 ↗
- Adversarial Poetry Refusal (AHB self-run) ↗
- Adversarial Poetry — AILuminate Baseline and Poetry ASR ↗
- Adversarial Robustness ↗
- Agent-SafetyBench ↗
- AILuminate General Purpose AI Chat ↗
- AIRBench 2024 Safety Scenarios ↗
- Anthropic Claude 4 System Card ↗
- Anthropic Claude Haiku 4.5 System Card ↗
- Anthropic Claude Opus 4.1 System Card Addendum ↗
- Anthropic Claude Opus 4.5 System Card ↗
- Anthropic Claude Sonnet 4.5 System Card ↗
- BioSecBench-Refusal V2 ↗
- BioTIER ↗
- BlueBench AttaQ-100 ↗
- CAIS Risk Index ↗
- ChiSafetyBench ↗
- Cisco AI Defense Rolling Single-Turn Leaderboard ↗
- Claude 2 model-card safety and alignment evaluations ↗
- Claude 3 model-card adversarial human-preference evaluations ↗
- Claude 3.5 Sonnet model-card safety and alignment evaluations ↗
- Claude Fable 5.1 / Mythos 5.1 card — safety and refusal calibration ↗
- Claude Opus 5.5 card — safety and refusal calibration ↗
- Claude Sonnet 4.6 Overrefusal ↗
- Claude Sonnet 4.6 User Wellbeing ↗
- COMPL-AI LLM RuLES Multi-Turn Rule Following ↗
- Concordia — AirBench-Deception ↗
- Concordia — AirBench-Manipulation ↗
- Concordia — AirBench-PoliticalPersuasion ↗
- Concordia — AirBench-SecurityRisks ↗
- Concordia — Fortress-Biological ↗
- Concordia — Fortress-Chemical ↗
- Concordia — Fortress-Privacy/Scams ↗
- Concordia — FRT-AirBench-Manipulation ↗
- Concordia — FRT-AirBench-SecurityRisks ↗
- Concordia — FRT-SciKnowEval-BiologicalHarmfulQA ↗
- Concordia — FRT-SOSBench-Chem ↗
- Concordia — SciKnowEval-BiologicalHarmfulQA ↗
- Concordia — SciKnowEval-ChemicalHarmfulQA ↗
- Concordia — SOSBench-Bio ↗
- Concordia — SOSBench-Chem ↗
- DecodingTrust ↗
- DelusionEval ↗
- Do-Not-Answer ↗
- DSPSafeBench ↗
- Enkrypt AI Safety Leaderboard ↗
- Fake Alignment (FINE) ↗
- FLAMES ↗
- FORTRESS ↗
- Google Gemini 2.5 Flash Model Card ↗
- Google Gemini 2.5 Flash-Lite Model Card ↗
- GPT-5.6 system card ↗
- GPT-6 Sol/Luna system card — safety and updated alignment tests ↗
- HarmBench ↗
- HELM Classic RealToxicityPrompts ↗
- HELM Safety ↗
- Human Pathogen Capabilities Test (HPCT) — overall refusal ↗
- HyperCLOVA X Toxic Continuation Panels ↗
- Inkling-Small model card — FORTRESS ↗
- Inkling-Small model card — StrongREJECT ↗
- JailBench ↗
- kindbench v0.1.0 psychological safety ranking ↗
- LiveSecBench ↗
- MedSafetyBench ↗
- Microsoft Phi Safety Panels ↗
- MT-JailBench CrescendoX ↗
- MuPPET Contextual Privacy ↗
- Open LLM Safety Index ↗
- OpenAI GPT-4o System Card ↗
- OpenAI GPT-5 System Card ↗
- OpenAI GPT-5.3 Dynamic Wellbeing ↗
- OpenAI GPT-5.4 Dynamic Wellbeing ↗
- OpenAI o3 and o4-mini System Card ↗
- OpenAI o3-mini System Card ↗
- OR-Bench ↗
- PandaBench JBB direct-request panel ↗
- PHARE ↗
- Qwen2 Safety Panel ↗
- S-Eval ↗
- SafeDialBench ↗
- SafetyPrompts ↗
- SALAD-Bench ↗
- Shell ↗
- SM-Bench ↗
- SORRY-Bench ↗
- SOSBench ↗
- SpeechMap model completion ↗
- SuperCLUE Safety ↗
- ThaiSafetyBench ↗
- TukaBench ↗
- UGI Leaderboard — base-model willingness ↗
- VETO Misfired Alignment ↗
- Vigil Mental Health Safety ↗
- XSTest ↗














