Models evaluated
637
Build 17 Nov 2026
As of
Canonical registry
Every model evaluated in the current snapshot, summarized by latent cross-board capability score, independent evidence coverage grade, and Pareto cost-frontier count.
637
Build 17 Nov 2026
71
Strong independent multi-source coverage
186
451 proprietary or unspecified
| Model | Consensus score | Evidence | Boards | Frontier | Pattern |
|---|---|---|---|---|---|
Gemini 4 CarbonGrade A. Evidence grade A of A to D: strongest. Google10 Nov 202618 boards | 98 pct: 100.0 (scale 0.0 to 100.0), 95% confidence interval 93.4 to 100.098 pct100.0(93.4–100.0) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Fable 5.2Grade A. Evidence grade A of A to D: strongest. Anthropic9 Nov 202615 boards | 99 pct: 95.3 (scale 0.0 to 100.0), 95% confidence interval 83.7 to 100.099 pct95.3(83.7–100.0) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Gemini 4 ArgonGrade A. Evidence grade A of A to D: strongest. Google13 Oct 202619 boards | 92 pct: 92.8 (scale 0.0 to 100.0), 95% confidence interval 88.1 to 95.592 pct92.8(88.1–95.5) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Opus 5.5Grade A. Evidence grade A of A to D: strongest. Anthropic22 Sept 202623 boards | 92 pct: 92.4 (scale 0.0 to 100.0), 95% confidence interval 84.8 to 99.092 pct92.4(84.8–99.0) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Sonnet 5.5Grade A. Evidence grade A of A to D: strongest. Anthropic28 Sept 202622 boards | 92 pct: 91.6 (scale 0.0 to 100.0), 95% confidence interval 83.7 to 96.892 pct91.6(83.7–96.8) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Gemini 4 HydrogenGrade A. Evidence grade A of A to D: strongest. Google12 Nov 202615 boards | 89 pct: 91.1 (scale 0.0 to 100.0), 95% confidence interval 84.1 to 97.389 pct91.1(84.1–97.3) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Muse Wonder Full 1.0Grade A. Evidence grade A of A to D: strongest. Meta12 Oct 202612 boards | 84 pct: 90.9 (scale 0.0 to 100.0), 95% confidence interval 85.0 to 94.684 pct90.9(85.0–94.6) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Fable 5.1Grade A. Evidence grade A of A to D: strongest. Anthropic1 Sept 202616 boards | 92 pct: 89.8 (scale 0.0 to 100.0), 95% confidence interval 81.4 to 94.292 pct89.8(81.4–94.2) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Muse Wonder 1.0Grade A. Evidence grade A of A to D: strongest. Meta12 Oct 202616 boards | 76 pct: 85.0 (scale 0.0 to 100.0), 95% confidence interval 78.7 to 88.376 pct85.0(78.7–88.3) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Fable 5Grade A. Evidence grade A of A to D: strongest. Anthropic9 Jun 202615 boards | 92 pct: 84.2 (scale 0.0 to 100.0), 95% confidence interval 76.6 to 90.892 pct84.2(76.6–90.8) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
GPT-5.5 Pre-releaseGrade C. Evidence grade C of A to D: thin. OpenAI23 Apr 20263 boards | 97 pct: 83.7 (scale 0.0 to 100.0), 95% confidence interval 72.0 to 89.797 pct83.7(72.0–89.7) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
GPT-6 AstraGrade A. Evidence grade A of A to D: strongest. OpenAI3 Sept 202625 boards | 86 pct: 83.5 (scale 0.0 to 100.0), 95% confidence interval 74.5 to 90.286 pct83.5(74.5–90.2) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
GPT-5.6 SolGrade A. Evidence grade A of A to D: strongest. OpenAI9 Jul 202616 boards | 90 pct: 83.5 (scale 0.0 to 100.0), 95% confidence interval 76.6 to 90.390 pct83.5(76.6–90.3) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Opus 5Grade A. Evidence grade A of A to D: strongest. Anthropic24 Jul 202616 boards | 92 pct: 83.3 (scale 0.0 to 100.0), 95% confidence interval 77.2 to 91.292 pct83.3(77.2–91.2) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
GPT-6.1 SolGrade A. Evidence grade A of A to D: strongest. OpenAI29 Sept 202622 boards | 85 pct: 83.1 (scale 0.0 to 100.0), 95% confidence interval 73.3 to 89.685 pct83.1(73.3–89.6) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Qwen4 MaxGrade C. Evidence grade C of A to D: thin. Alibaba9 boards | 73 pct: 82.7 (scale 0.0 to 100.0), 95% confidence interval 76.0 to 89.173 pct82.7(76.0–89.1) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
DeepSeek V4.1 Flash 1017Grade A. Evidence grade A of A to D: strongest. DeepSeekOpen weights17 Oct 202615 boards | 68 pct: 82.4 (scale 0.0 to 100.0), 95% confidence interval 75.5 to 86.868 pct82.4(75.5–86.8) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Muse EmberGrade B. Evidence grade B of A to D: good. Meta9 Nov 202611 boards | 67 pct: 81.5 (scale 0.0 to 100.0), 95% confidence interval 73.8 to 87.467 pct81.5(73.8–87.4) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
GPT-5.4 ProGrade C. Evidence grade C of A to D: thin. OpenAI5 Mar 20264 boards | 93 pct: 81.3 (scale 0.0 to 100.0), 95% confidence interval 71.5 to 85.393 pct81.3(71.5–85.3) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
Opus 4.7Grade B. Evidence grade B of A to D: good. Anthropic16 Apr 202617 boards | 90 pct: 80.3 (scale 0.0 to 100.0), 95% confidence interval 73.8 to 87.490 pct80.3(73.8–87.4) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
Opus 4.6Grade A. Evidence grade A of A to D: strongest. Anthropic5 Feb 202618 boards | 91 pct: 80.3 (scale 0.0 to 100.0), 95% confidence interval 71.7 to 87.491 pct80.3(71.7–87.4) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
GPT-5.5Grade B. Evidence grade B of A to D: good. OpenAI23 Apr 202614 boards | 87 pct: 79.9 (scale 0.0 to 100.0), 95% confidence interval 74.3 to 84.887 pct79.9(74.3–84.8) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
Qwen3.8-MaxGrade A. Evidence grade A of A to D: strongest. Alibaba2 Aug 202610 boards | 92 pct: 79.1 (scale 0.0 to 100.0), 95% confidence interval 69.9 to 87.792 pct79.1(69.9–87.7) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Kimi K3.1 ProGrade A. Evidence grade A of A to D: strongest. Moonshot AIOpen weights23 Oct 202615 boards | 70 pct: 78.5 (scale 0.0 to 100.0), 95% confidence interval 69.5 to 84.270 pct78.5(69.5–84.2) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Within-domain dispute | ||
Muse Spark 1.3Grade B. Evidence grade B of A to D: good. Meta2 Sept 202610 boards | 79 pct: 78.2 (scale 0.0 to 100.0), 95% confidence interval 72.4 to 84.779 pct78.2(72.4–84.7) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
Qwen3.8 Max InternalGrade C. Evidence grade C of A to D: thin. Alibaba9 boards | 59 pct: 77.8 (scale 0.0 to 100.0), 95% confidence interval 64.5 to 86.659 pct77.8(64.5–86.6) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
GPT-6 SolGrade A. Evidence grade A of A to D: strongest. OpenAI22 Sept 202613 boards | 82 pct: 76.8 (scale 0.0 to 100.0), 95% confidence interval 69.5 to 84.982 pct76.8(69.5–84.9) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Grok 4.7Grade A. Evidence grade A of A to D: strongest. xAI21 Sept 202621 boards | 68 pct: 76.4 (scale 0.0 to 100.0), 95% confidence interval 70.6 to 80.568 pct76.4(70.6–80.5) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
GPT-6.1 TerraGrade A. Evidence grade A of A to D: strongest. OpenAI26 Oct 202614 boards | 75 pct: 76.1 (scale 0.0 to 100.0), 95% confidence interval 63.8 to 86.275 pct76.1(63.8–86.2) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Within-domain dispute | ||
Hy4Grade B. Evidence grade B of A to D: good. Tencent3 Nov 202610 boards | 59 pct: 75.7 (scale 0.0 to 100.0), 95% confidence interval 66.6 to 81.259 pct75.7(66.6–81.2) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
GLM-5.3Grade A. Evidence grade A of A to D: strongest. Z.aiOpen weights14 Aug 202618 boards | 64 pct: 75.6 (scale 0.0 to 100.0), 95% confidence interval 69.0 to 79.964 pct75.6(69.0–79.9) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Within-domain dispute | ||
Gemini 3.7 FlashGrade A. Evidence grade A of A to D: strongest. Google13 Aug 202612 boards | 82 pct: 75.6 (scale 0.0 to 100.0), 95% confidence interval 65.9 to 82.182 pct75.6(65.9–82.1) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Domain specialist | ||
Inkling 2 SmallGrade A. Evidence grade A of A to D: strongest. Thinking Machines5 Nov 202614 boards | 68 pct: 74.8 (scale 0.0 to 100.0), 95% confidence interval 65.3 to 84.968 pct74.8(65.3–84.9) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Within-domain dispute | ||
MiMo-V2.6-ProGrade A. Evidence grade A of A to D: strongest. Xiaomi21 Sept 202616 boards | 65 pct: 74.4 (scale 0.0 to 100.0), 95% confidence interval 67.1 to 79.965 pct74.4(67.1–79.9) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Gemini 3.8 FlashGrade A. Evidence grade A of A to D: strongest. Google2 Sept 202623 boards | 64 pct: 73.8 (scale 0.0 to 100.0), 95% confidence interval 65.4 to 80.264 pct73.8(65.4–80.2) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Within-domain dispute | ||
Muse SparkGrade B. Evidence grade B of A to D: good. Meta8 Apr 20269 boards | 89 pct: 73.3 (scale 0.0 to 100.0), 95% confidence interval 63.3 to 81.689 pct73.3(63.3–81.6) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
GPT-5.6 TerraGrade A. Evidence grade A of A to D: strongest. OpenAI9 Jul 202614 boards | 74 pct: 73.0 (scale 0.0 to 100.0), 95% confidence interval 66.8 to 78.874 pct73.0(66.8–78.8) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Gemini 3.5 FlashGrade B. Evidence grade B of A to D: good. Google19 May 202612 boards | 87 pct: 72.8 (scale 0.0 to 100.0), 95% confidence interval 63.9 to 78.687 pct72.8(63.9–78.6) | Grade B. Evidence grade B of A to D: good.2 indep. | Within-domain dispute | ||
Muse Spark 1.1Grade B. Evidence grade B of A to D: good. Meta9 Jul 20269 boards | 79 pct: 72.7 (scale 0.0 to 100.0), 95% confidence interval 63.5 to 79.379 pct72.7(63.5–79.3) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
GPT-5.4Grade A. Evidence grade A of A to D: strongest. OpenAI5 Mar 202617 boards | 77 pct: 72.6 (scale 0.0 to 100.0), 95% confidence interval 65.2 to 77.577 pct72.6(65.2–77.5) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Kimi K3Grade A. Evidence grade A of A to D: strongest. Moonshot AIOpen weights16 Jul 202618 boards | 61 pct: 72.3 (scale 0.0 to 100.0), 95% confidence interval 64.2 to 78.161 pct72.3(64.2–78.1) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
Haiku 5.5Grade A. Evidence grade A of A to D: strongest. Anthropic7 Oct 202620 boards | 56 pct: 71.9 (scale 0.0 to 100.0), 95% confidence interval 62.7 to 76.356 pct71.9(62.7–76.3) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Consistent | ||
DeepSeek V4.1 FlashGrade A. Evidence grade A of A to D: strongest. DeepSeekOpen weights9 Sept 20269 boards | 78 pct: 71.6 (scale 0.0 to 100.0), 95% confidence interval 65.5 to 76.478 pct71.6(65.5–76.4) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Gemini 3.1 Pro PreviewGrade A. Evidence grade A of A to D: strongest. Google19 Feb 202617 boards | 83 pct: 71.5 (scale 0.0 to 100.0), 95% confidence interval 61.8 to 77.783 pct71.5(61.8–77.7) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Within-domain dispute | ||
Grok 4.6Grade A. Evidence grade A of A to D: strongest. xAI12 Aug 202614 boards | 76 pct: 70.7 (scale 0.0 to 100.0), 95% confidence interval 64.2 to 75.976 pct70.7(64.2–75.9) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Within-domain dispute | ||
Opus 4.8Grade A. Evidence grade A of A to D: strongest. Anthropic28 May 202615 boards | 74 pct: 70.5 (scale 0.0 to 100.0), 95% confidence interval 60.1 to 78.874 pct70.5(60.1–78.8) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Mistral Large 4 OSSGrade B. Evidence grade B of A to D: good. MistralOpen weights29 Oct 202610 boards | 43 pct: 70.2 (scale 0.0 to 100.0), 95% confidence interval 61.9 to 79.643 pct70.2(61.9–79.6) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
GLM-5.2Grade B. Evidence grade B of A to D: good. Z.aiOpen weights16 Jun 20269 boards | 79 pct: 70.1 (scale 0.0 to 100.0), 95% confidence interval 61.3 to 78.679 pct70.1(61.3–78.6) | Grade B. Evidence grade B of A to D: good.2 indep. | Within-domain dispute | ||
Sonnet 4.6Grade B. Evidence grade B of A to D: good. Anthropic17 Feb 202617 boards | 73 pct: 70.0 (scale 0.0 to 100.0), 95% confidence interval 63.8 to 75.373 pct70.0(63.8–75.3) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
Gemini 3.6 FlashGrade B. Evidence grade B of A to D: good. Google21 Jul 202610 boards | 74 pct: 70.0 (scale 0.0 to 100.0), 95% confidence interval 59.5 to 76.874 pct70.0(59.5–76.8) | Grade B. Evidence grade B of A to D: good.2 indep. | Within-domain dispute | ||
MiMo-V2.6-FlashGrade A. Evidence grade A of A to D: strongest. Xiaomi21 Sept 202616 boards | 45 pct: 68.8 (scale 0.0 to 100.0), 95% confidence interval 62.1 to 74.445 pct68.8(62.1–74.4) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Domain specialist | ||
ERNIE 5.1Grade C. Evidence grade C of A to D: thin. Baidu4 boards | 82 pct: 68.1 (scale 0.0 to 100.0), 95% confidence interval 52.7 to 76.582 pct68.1(52.7–76.5) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
Kimi K2.6Grade B. Evidence grade B of A to D: good. Moonshot AIOpen weights20 Apr 202611 boards | 79 pct: 67.8 (scale 0.0 to 100.0), 95% confidence interval 60.0 to 72.579 pct67.8(60.0–72.5) | Grade B. Evidence grade B of A to D: good.2 indep. | Consistent | ||
Hy4 PreviewGrade A. Evidence grade A of A to D: strongest. Tencent12 boards | 35 pct: 67.7 (scale 0.0 to 100.0), 95% confidence interval 58.9 to 74.135 pct67.7(58.9–74.1) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
Grok 4.5Grade A. Evidence grade A of A to D: strongest. xAI8 Jul 202614 boards | 64 pct: 67.3 (scale 0.0 to 100.0), 95% confidence interval 59.5 to 73.264 pct67.3(59.5–73.2) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Within-domain dispute | ||
GPT-5.2 ChatGrade C. Evidence grade C of A to D: thin. OpenAI4 boards | 83 pct: 67.1 (scale 0.0 to 100.0), 95% confidence interval 55.6 to 73.483 pct67.1(55.6–73.4) | Grade C. Evidence grade C of A to D: thin.1 indep. | Consistent | ||
GLM-5.3 FlashGrade A. Evidence grade A of A to D: strongest. Z.aiOpen weights20 Aug 202620 boards | 43 pct: 66.9 (scale 0.0 to 100.0), 95% confidence interval 59.0 to 70.843 pct66.9(59.0–70.8) | Grade A. Evidence grade A of A to D: strongest.4 indep. | Within-domain dispute | ||
GPT-5.3 CodexGrade A. Evidence grade A of A to D: strongest. OpenAI5 Feb 20265 boards | 64 pct: 66.8 (scale 0.0 to 100.0), 95% confidence interval 56.6 to 73.864 pct66.8(56.6–73.8) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
GPT-5.6 LunaGrade A. Evidence grade A of A to D: strongest. OpenAI9 Jul 202613 boards | 61 pct: 66.7 (scale 0.0 to 100.0), 95% confidence interval 59.7 to 72.261 pct66.7(59.7–72.2) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Consistent | ||
DeepSeek V4 Pro 0813Grade A. Evidence grade A of A to D: strongest. DeepSeekOpen weights13 Aug 202615 boards | 43 pct: 66.6 (scale 0.0 to 100.0), 95% confidence interval 58.7 to 71.543 pct66.6(58.7–71.5) | Grade A. Evidence grade A of A to D: strongest.3 indep. | Within-domain dispute |