Evaluation Data

Currently showing：Run #154 | 2026-06-08 | 212-question pool | Formula v7 | Judge set v6.1

Switch Run Run #168 — Full — 06-13 Run #166 — Smoke — 06-13 Run #165 — Smoke — 06-12 Run #163 — Full — 06-11 Run #162 — Full — 06-11 Run #160 — Full — 06-11 Run #159 — Smoke — 06-11 Run #158 — Smoke — 06-11 Run #156 — Smoke — 06-10 Run #155 — Smoke — 06-09 Run #154 — Full — 06-08 (Current) Run #153 — Smoke — 06-08 Run #152 — Smoke — 06-07 Run #151 — Smoke — 06-06 Run #150 — targeted_smoke — 06-06 Run #149 — Smoke — 06-06 Run #148 — Smoke — 06-05 Run #147 — Smoke — 06-04 Run #145 — Smoke — 06-03 Run #144 — targeted_smoke — 06-02 Run #143 — Smoke — 06-02 Run #142 — Full — 06-01 Run #141 — Smoke — 06-01 Run #139 — Smoke — 05-31 Run #138 — Smoke — 05-30 Run #137 — Smoke — 05-29 Run #136 — Smoke — 05-28 Run #134 — Smoke — 05-27 Run #133 — targeted_smoke — 05-26 Run #132 — Smoke — 05-26 Run #131 — Full — 05-25 Run #130 — Smoke — 05-25 Run #129 — Smoke — 05-24 Run #128 — Smoke — 05-23 Run #127 — Smoke — 05-22 Run #126 — Smoke — 05-21 Run #124 — Smoke — 05-20 Run #123 — Smoke — 05-19 Run #122 — Full — 05-18 Run #121 — Smoke — 05-18 Run #119 — Smoke — 05-17 Run #118 — Smoke — 05-16 Run #117 — Smoke — 05-15 Run #116 — Smoke — 05-14 Run #114 — Smoke — 05-13 Run #113 — Smoke — 05-12 Run #112 — Full — 05-11 Run #111 — Smoke — 05-11 Run #110 — targeted_smoke — 05-10 Run #109 — Smoke — 05-10 Run #108 — Smoke — 05-09 Run #107 — Smoke — 05-08 Run #106 — Smoke — 05-07 Run #104 — Smoke — 05-06 Run #103 — Smoke — 05-05 Run #102 — Full — 05-04 Run #101 — Smoke — 05-04 Run #99 — Smoke — 05-03 Run #98 — Smoke — 05-02 Run #91 — Smoke — 05-01 Run #90 — Smoke — 04-30 Run #89 — Smoke — 04-29 Run #88 — Smoke — 04-28 Run #87 — Full — 04-27 Run #86 — Smoke — 04-27 Run #85 — Smoke — 04-26 Run #84 — Smoke — 04-25 Run #83 — Smoke — 04-24 Run #82 — Smoke — 04-23 Run #81 — Smoke — 04-22 Run #80 — Smoke — 04-21 Run #79 — Smoke — 04-21 Run #78 — Full — 04-20 Run #77 — Smoke — 04-20 Run #76 — Smoke — 04-19 Run #75 — Smoke — 04-18 Run #73 — Smoke — 04-17 Run #72 — Smoke — 04-16 Run #71 — Smoke — 04-15 Run #70 — Smoke — 04-14 Run #69 — Full — 04-13 Run #68 — Smoke — 04-13 Run #67 — Smoke — 04-12 Run #66 — Smoke — 04-11 Run #65 — Smoke — 04-10 Run #64 — Smoke — 04-09 Run #63 — Smoke — 04-08 Run #62 — Smoke — 04-07 Run #61 — Full — 04-06 Run #60 — Smoke — 04-06 Run #59 — Smoke — 04-05 Run #58 — targeted_smoke — 04-04 Run #57 — Smoke — 04-04 Run #56 — Smoke — 04-03 Run #55 — Smoke — 04-02 Run #54 — Smoke — 04-01 Run #53 — Smoke — 03-31 Run #52 — Full — 03-30 Run #51 — targeted_smoke — 03-30 Run #50 — Smoke — 03-30 Run #49 — Smoke — 03-29 Run #47 — Smoke — 03-28 Run #46 — Smoke — 03-27 Run #43 — Full — 03-25 Run #42 — Smoke — 03-25 Run #41 — Full — 03-24 Run #40 — Full — 03-24 Run #39 — Full — 03-24 Run #38 — Full — 03-24 Run #37 — Full — 03-22 Run #36 — Smoke — 03-22 Run #35 — Full — 03-22 Run #33 — Full — 03-21 Run #32 — Smoke — 03-21 Run #31 — Full — 03-21 Run #30 — Full — 03-21 Run #29 — Full — 03-21 Run #27 — Full — 03-21 Run #26 — Smoke — 03-21 Run #25 — Smoke — 03-21 Run #24 — Smoke — 03-21 Run #23 — Smoke — 03-20 Run #22 — Smoke — 03-20 Run #20 — Full — 03-19 Run #18 — Smoke — 03-19 Run #17 — Smoke — 03-18 Run #16 — Full — 03-18 Run #15 — Full — 03-17 Run #14 — Full — 03-17 Run #12 — Smoke — 03-17 Run #11 — Full — 03-17 Run #10 — Full — 03-17 Run #9 — Full — 03-17 Run #8 — Smoke — 03-16 Run #7 — Smoke — 03-16 Run #6 — Smoke — 03-16 Run #5 — Smoke — 03-16 Run #4 — Smoke — 03-16 Run #3 — Smoke — 03-16 Run #2 — Smoke — 03-16 Run #1 — Smoke — 03-16 Model All (11) Claude Opus 4.7 Claude Sonnet 4.6 DeepSeek V4 Pro Gemini 2.5 Pro Gemini 3.1 Pro GPT-5.5 GPT-o3 Grok 4 Qwen3 Max ERNIE Bot 4.5 Doubao Pro

Data Disclosure：To prevent benchmark contamination and overfitting, question texts and expected answers are not disclosed. This page shows model responses, scores, and judging methods for transparency. For the full methodology, seeMethodology page。

Model	Code Execution	Grounding	Engineering Judgment	Task Communication	Integrity	Overallpts	Value	Stability	Availability	Per Task
Grok 4 grok	93.90	85.00	82.10	87.80	86.30 pass	89.90	29.7	68.6	100.0	Expand
Claude Opus 4.7 claude	90.30	87.50	93.10	89.40	94.30 pass	89.04	6.2	67.7	100.0	Expand
Doubao Pro doubao	94.60	81.60	88.80	84.10	92.20 pass	88.75	96.2	71.2	100.0	Expand
Claude Sonnet 4.6 claude	87.60	86.80	93.20	87.80	94.70 pass	87.24	29.7	62.7	100.0	Expand
Gemini 2.5 Pro gemini	88.10	84.20	87.70	84.60	88.80 pass	86.35	44.6	66.0	99.0	Expand
Qwen3 Max qwen	89.70	81.90	85.70	85.30	87.50 pass	86.19	58.5	59.8	100.0	Expand
Gemini 3.1 Pro gemini	88.40	80.40	85.20	84.90	87.70 pass	84.80	29.3	63.2	99.0	Expand
DeepSeek V4 Pro deepseek	87.90	77.60	82.40	85.10	81.80 pass	83.27	47.5	59.1	100.0	Expand
GPT-o3 gpt	84.80	80.40	91.50	87.50	90.60 pass	82.82	10.5	58.0	100.0	Expand
GPT-5.5 gpt	81.90	79.70	92.10	87.40	88.30 pass	80.91	20.4	51.8	100.0	Expand
ERNIE Bot 4.5 ernie	78.00	75.60	72.20	72.00	70.00 pass	76.92	99.3	44.2	100.0	Expand

API Access：For programmatic access to evaluation data, please use our API。