A benchmark for OpenClaw agents
OpenClawProBench
A benchmark designed to measure how different models perform under OpenClaw when they need to reason, plan, use tools, and stay reliable across repeated runs.
Benchmark Leaderboard
Pass^3 remains the default rank signal. The table below shows score quality, efficiency, all six evaluation dimensions, average runtime, total token usage, token breakdown, OpenClaw version, and update time.
Choose Columns
Current sorting will be preserved.
| # | Model | Pass^3 | Pass@3 | Avg Score | Capability | Efficiency | Planning | Safety | Tool Use | Constraints | Error Recovery | Synthesis | Avg Runtime | Total Token | Token | Cost | Price | OpenClaw | Released | Updated |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 |
|
49.6%
|
61.2 | 70.1 | 70.9 | 91.1 | 76.5 | 67.2 | 70.7 | 69.6 | 75.0 | 59.4 | 123.72s | 74256276 |
input:10347477
output:534393
cacheread:63374406
|
$11.7627 |
$0.26 / $1.56
|
v3.2.6 | 2026-02-15 | 2026-04-06 |
| 2 |
|
44.9%
|
61.6 | 69.0 | 69.1 | 98.8 | 75.2 | 68.4 | 71.0 | 65.8 | 74.1 | 56.7 | 89.33s | 47621722 |
input:2299534
output:356684
cacheread:44965504
|
$25.9237 |
$1 / $3.2
¥6.88 / ¥22.02
|
v3.2.6 | 2026-03-27 | 2026-03-30 |
| 3 |
|
41.6%
|
57.5 | 68.3 | 68.6 | 97.4 | 70.1 | 68.3 | 73.1 | 65.1 | 73.8 | 57.4 | 90.89s | 20474370 |
input:3405226
output:188658
cacheread:16880486
|
$17.8378 |
$1.3948 / $6.9741
¥9.6 / ¥48.0
|
v3.2.6 | 2026-02-14 | 2026-04-04 |
| 4 |
|
43.2%
|
62.9 | 67.7 | 68.6 | 90.3 | 70.3 | 67.5 | 69.1 | 62.9 | 74.6 | 60.7 | 149.96s | 33552869 |
input:3918340
output:191987
cacheread:29442542
|
$27.3375 |
$1.3948 / $6.9741
¥9.6 / ¥48.0
|
v3.2.6 | 2026-02-14 | 2026-04-04 |
| 5 |
|
39.0%
|
57.0 | 67.6 | 68.4 | 90.6 | 68.7 | 65.8 | 68.9 | 65.3 | 73.3 | 62.8 | 160.91s | 65468885 |
input:10004682
output:637259
cacheread:54826944
|
$9.9709 |
$0.26 / $0.38
|
v3.2.6 | 2025-12-01 | 2026-04-03 |
| 6 |
|
46.9%
|
55.0 | 67.1 | 67.4 | 96.1 | 72.2 | 66.6 | 67.4 | 65.3 | 70.5 | 58.6 | 80.31s | 27059691 |
input:2273484
output:326612
cacheread:24459595
|
$18.7104 |
$1.2 / $4
|
v3.2.6 | 2026-03-16 | 2026-04-04 |
| 7 |
|
36.9%
|
60.8 | 66.7 | 68.0 | 86.9 | 68.8 | 67.1 | 64.2 | 65.4 | 72.5 | 62.5 | 201.7s | 97734852 |
input:10027746
output:545655
cacheread:87161451
|
$14.1456 |
$0.26 / $0.38
|
v3.2.6 | 2025-12-01 | 2026-04-06 |
| 8 |
|
45.5%
|
53.9 | 66.6 | 67.5 | 90.9 | 71.6 | 65.0 | 64.0 | 66.8 | 71.4 | 60.2 | 178.58s | 29132003 |
input:28539688
output:592315
|
$94.5038 |
$3 / $15
|
v3.2.6 | 2026-02-18 | 2026-04-06 |
| 9 |
|
30.7%
|
60.3 | 65.5 | 66.8 | 87.4 | 68.0 | 67.4 | 61.5 | 62.6 | 71.4 | 62.7 | 160.53s | 72630744 |
input:71928542
output:702202
|
$18.9683 |
$0.26 / $0.38
|
v3.2.6 | 2025-12-01 | 2026-04-06 |
| 10 |
|
33.9%
|
59.0 | 65.4 | 65.9 | 93.8 | 68.3 | 64.8 | 66.9 | 61.2 | 72.1 | 57.4 | 88.73s | 41679573 |
input:3470897
output:369572
cacheread:37839104
|
$9.2045 |
$0.3827 / $1.72
|
v3.2.6 | 2026-01-27 | 2026-04-02 |
| 11 |
|
39.0%
|
55.9 | 65.3 | 66.0 | 91.2 | 71.2 | 63.3 | 64.1 | 66.7 | 68.4 | 56.2 | 120.83s | 62909389 |
input:10724289
output:627594
cacheread:51557506
|
$46.7179 |
$1.16 / $6.97
¥8 / ¥48
|
v3.2.6 | 2026-04-02 | 2026-04-04 |
| 12 |
|
37.6%
|
53.2 | 64.7 | 65.1 | 95.9 | 70.3 | 64.1 | 60.1 | 66.2 | 66.5 | 60.3 | 107.16s | 46485163 |
input:12390088
output:665699
cacheread:33429376
|
$31.1019 |
$1 / $3
|
v3.2.6 | 2026-03-19 | 2026-04-01 |
| 13 |
|
34.8%
|
54.6 | 64.5 | 65.0 | 94.9 | 69.4 | 63.0 | 60.2 | 65.4 | 70.3 | 58.3 | 205.9s | 40976389 |
input:40405158
output:571231
|
$0 |
$0 / $0
¥0 / ¥0
|
v3.2.6 | 2026-01-16 | 2026-04-04 |
| 14 |
|
28.5%
|
61.8 | 64.0 | 64.5 | 95.9 | 69.7 | 65.5 | 62.5 | 59.3 | 71.9 | 53.6 | 122.49s | 34977005 |
input:3229307
output:564274
cacheread:31183424
|
$14.849 |
$0.72 / $2.3
|
v3.2.6 | 2026-02-11 | 2026-03-31 |
| 15 |
|
29.5%
|
57.0 | 63.6 | 64.3 | 93.4 | 64.8 | 65.2 | 64.8 | 59.6 | 69.7 | 57.0 | 104.98s | 42991923 |
input:2767151
output:872645
cacheread:39352127
|
$10.4109 |
$0.39 / $1.9
|
v3.2.6 | 2025-09-30 | 2026-04-05 |
| 16 |
|
35.2%
|
56.6 | 63.5 | 64.0 | 93.1 | 65.4 | 65.6 | 63.5 | 59.8 | 68.9 | 57.2 | 74.53s | 40532601 |
input:12022794
output:801205
cacheread:27708602
|
$11.9532 |
$0.4 / $2
|
v3.2.6 | 2026-03-19 | 2026-04-01 |
| 17 |
|
32.8%
|
55.1 | 63.4 | 64.9 | 84.6 | 66.6 | 63.2 | 64.3 | 59.1 | 66.3 | 59.8 | 125.1s | 48417591 |
input:36403992
output:507295
cacheread:11506304
|
$113.0023 |
$2.5 / $15
|
v3.2.6 | 2026-03-05 | 2026-04-02 |
| 18 |
|
26.4%
|
54.8 | 62.7 | 63.3 | 94.3 | 65.1 | 63.9 | 60.5 | 63.8 | 69.2 | 53.6 | 124.51s | 40323216 |
input:3276264
output:829547
cacheread:36217405
|
$9.7918 |
$0.39 / $1.75
|
v3.2.6 | 2025-12-23 | 2026-04-01 |
| 19 |
|
31.7%
|
52.3 | 60.3 | 61.3 | 88.7 | 63.5 | 63.0 | 62.2 | 60.7 | 63.3 | 47.7 | 107.08s | 98806785 |
input:16915312
output:544641
cacheread:81346832
|
$17.9302 |
$0.3 / $1.2
|
v3.2.6 | 2026-03-27 | 2026-04-01 |
| 20 |
|
28.3%
|
47.2 | 60.1 | 60.9 | 91.8 | 57.8 | 65.4 | 57.7 | 59.8 | 69.5 | 52.1 | 162.28s | 54383209 |
input:53285801
output:1097408
|
$44.4101 |
$0.77 / $3.08
¥5.3 / ¥21.2
|
v3.2.6 | 2025-12-05 | 2026-04-04 |
| 21 |
|
36.6%
|
57.8 | 59.8 | 60.2 | 95.5 | 72.8 | 64.1 | 62.9 | 57.6 | 58.7 | 37.2 | 220.61s | 67262104 |
input:11956034
output:284269
cacheread:55021801
|
$41.2943 |
$1.017 / $4.068
¥7 / ¥28
|
v3.2.6 | 2026-01-23 | 2026-04-07 |
| 22 |
|
23.3%
|
52.7 | 58.8 | 59.4 | 92.6 | 57.6 | 63.1 | 61.1 | 60.5 | 60.8 | 49.0 | 96.4s | 54177427 |
input:8780828
output:669815
cacheread:44726784
|
$4.3381 |
$0.118 / $0.99
|
v3.2.6 | 2026-02-12 | 2026-04-02 |
| 23 |
|
30.1%
|
54.6 | 58.1 | 58.6 | 95.1 | 57.4 | 61.0 | 60.0 | 54.2 | 66.2 | 49.4 | 107.66s | 49678586 |
input:14860816
output:717196
cacheread:34100574
|
$72.4286 |
$2 / $12
|
v3.2.6 | 2026-02-20 | 2026-04-02 |
| 24 |
|
15.9%
|
45.1 | 57.1 | 58.1 | 90.8 | 57.8 | 62.1 | 56.3 | 55.6 | 61.9 | 48.9 | 122.05s | 63879796 |
input:63382367
output:497429
|
$41.9994 |
$0.65 / $1.61
¥4.5 / ¥11.1
|
v3.2.6 | 2025-12-05 | 2026-04-05 |
| 25 |
|
20.2%
|
50.0 | 56.8 | 57.3 | 94.8 | 59.0 | 60.0 | 52.3 | 55.4 | 63.6 | 51.5 | 134.51s | 70440558 |
input:8033465
output:678335
cacheread:59133252
cachewrite:2595506
|
$12.4834 |
$0.3 / $1.2
|
v3.2.6 | 2026-03-18 | 2026-04-02 |
| 26 |
|
16.4%
|
50.5 | 56.8 | 57.3 | 93.1 | 58.8 | 59.0 | 56.0 | 56.4 | 60.6 | 49.3 | 92.82s | 49828422 |
input:9501688
output:677502
cacheread:39649232
|
$8.5617 |
$0.27 / $0.95
|
v3.2.6 | 2025-12-23 | 2026-04-01 |
| 27 |
|
27.8%
|
50.2 | 54.6 | 55.5 | 89.3 | 56.0 | 55.6 | 60.5 | 57.8 | 52.5 | 42.9 | 508.52s | 63436733 |
input:8824540
output:336271
cacheread:54275922
|
$21.676 |
$0.581 / $2.325
¥4 / ¥16
|
v3.2.6 | 2026-02-16 | 2026-04-07 |
| 28 |
|
19.4%
|
41.8 | 54.6 | 55.7 | 87.7 | 56.4 | 59.3 | 56.5 | 52.3 | 57.7 | 43.9 | 71.83s | 104722641 |
input:12293170
output:381362
cacheread:92048109
|
$21.7349 |
$0.3632 / $1.4529
¥2.5 / ¥10
|
v3.2.6 | 2026-02-20 | 2026-04-08 |
| 29 |
|
22.4%
|
48.9 | 53.8 | 54.3 | 93.0 | 54.2 | 63.0 | 37.8 | 57.1 | 61.9 | 53.7 | 110.14s | 109050773 |
input:26076719
output:244800
cacheread:82729254
|
$34.4551 |
$0.5 / $3
|
v3.2.6 | 2025-12-17 | 2026-04-05 |
| 30 |
|
14.2%
|
48.9 | 52.4 | 53.0 | 90.7 | 50.2 | 57.0 | 48.0 | 51.7 | 58.6 | 50.8 | 109.15s | 43699451 |
input:18648766
output:1078701
cacheread:23971984
|
$5.2424 |
$0.15 / $0.6
|
v3.2.6 | 2026-03-16 | 2026-04-06 |
| 31 |
|
15.3%
|
32.9 | 52.3 | 53.6 | 86.9 | 50.1 | 56.9 | 55.4 | 50.9 | 58.4 | 41.9 | 107.1s | 86247300 |
input:85197273
output:1050027
|
$8.8452 |
$0.1 / $0.31
¥0.7 / ¥2.1
|
v3.2.6 | 2026-04-02 | 2026-04-02 |
| 32 |
|
14.7%
|
31.1 | 51.8 | 53.0 | 86.1 | 50.1 | 58.8 | 54.3 | 49.8 | 57.3 | 40.0 | 103.62s | 87855843 |
input:86724853
output:1130990
|
$9.0118 |
$0.1 / $0.3
|
v3.2.6 | 2026-02-02 | 2026-04-04 |
| 33 |
|
14.7%
|
42.1 | 48.8 | 49.6 | 92.3 | 60.2 | 44.7 | 43.4 | 48.7 | 50.4 | 43.5 | 80.19s | 51133010 |
input:7668306
output:197632
cacheread:43267072
|
$59.7895 |
$2 / $6
|
v3.2.6 | 2026-03-12 | 2026-04-04 |
| 34 |
|
14.7%
|
41.2 | 48.4 | 49.0 | 89.9 | 46.6 | 55.6 | 50.1 | 45.3 | 53.6 | 39.4 | 242.33s | 32609723 |
input:31554432
output:1055291
|
$14.3483 |
$0.44 / $0.44
¥3 / ¥3
|
v3.2.6 | 2026-02-11 | 2026-04-07 |
| 35 |
|
9.5%
|
35.3 | 47.9 | 48.7 | 91.2 | 42.9 | 53.7 | 54.7 | 39.4 | 55.6 | 40.2 | 82.69s | 63192773 |
input:13974967
output:1641102
cacheread:47576704
|
$4.2687 |
$0.1 / $0.3
|
v3.2.6 | 2026-02-02 | 2026-04-01 |
| 36 |
|
7.1%
|
26.0 | 42.9 | 43.0 | 96.6 | 29.1 | 52.1 | 44.0 | 40.8 | 54.6 | 41.0 | 115.4s | 32527395 |
input:9183126
output:467510
cacheread:22876759
|
$2.6896 |
$0.12 / $0.46
¥0.8 / ¥3.2
|
v3.2.6 | 2026-04-02 | 2026-04-05 |
| 11 |
|
29.4%
|
50.3 | 61.7 | 62.4 | 92.0 | 64.4 | 63.8 | 60.2 | 60.7 | 68.9 | 51.9 | 104.92s | 66378687 |
input:11027658
output:621286
cacheread:54729743
|
$48.9604 |
$1.1624 / $6.9741
¥8 / ¥48
|
v3.2.6 | 2026-04-02 | 2026-04-09 |
| 6 |
|
40.4%
|
62.9 | 65.0 | 65.8 | 93.4 | 63.6 | 67.2 | 68.9 | 59.8 | 72.7 | 56.7 | 105.51s | 79743715 |
input:10387510
output:569004
cacheread:68787201
|
$19.5397 |
$0.4068 / $2.3247
¥2.8 / ¥16.0
|
v3.2.6 | 2026-02-14 | 2026-04-09 |
Notes:
USD/CNY quick reference uses approximately
1($)≈ 6.8826(¥).
cache_read_tokens and cache_write_tokens are billed at 0.5x the input-token price.
cost_usd = (input_tokens + 0.5 * (cache_read_tokens + cache_write_tokens)) * price_usd_input / 1e6 + output_tokens * price_usd_output / 1e6
Visual Leaderboard
View the full leaderboard in chart form. When you switch metrics, the bar ranking below re-sorts by the selected field.
qwen3.5-plus
qwen · bailiancodingplan
GLM-5.1
glm · glm
doubao-seed-2.0-pro
seed · volcengine-plan
doubao-seed-2.0-code
seed · volcengine-plan
DeepSeek-V3.2
deepseek-ai · siliconflow
GLM-5-Turbo
glm · glm
DeepSeek-V3.2
deepseek-ai · volcengine-plan
Claude Sonnet 4.6
anthropic · openrouter
DeepSeekV3.2
deepseek-ai · baiduqianfan
kimi-k2.5
moonshot · moonshot
qwen3.6-plus
qwen · bailianapi
mimo-v2-pro
xiaomi · openrouter
LongCat-Flash-Thinking-2601
meituan · longcat
GLM-5
glm · glm
GLM-4.6
glm · glm
mimo-v2-omni
xiaomi · openrouter
gpt-5.4
openai · openrouter
GLM-4.7
glm · glm
kat-coder-pro-v2
kwaipilot · openrouter
hunyuan-2.0-thinking
tencent · tencent-token-plan
qwen3-max-2026-01-23
qwen · bailiancodingplan
MiniMax-M2.5
minimax · minimax
gemini-3.1-pro-preview
google · openrouter
hunyuan-2.0-instruct
tencent · tencent-token-plan
MiniMax-M2.7
minimax · minimax
MiniMax-M2.1
minimax · minimax
Ling-2.5-1T
tbox · antling
qwen3-coder-next
qwen · bailiancodingplan
gemini-3-flash-preview
google · openrouter
mistral-small-2603
mistralai · openrouter
step-3.5-flash-2603
stepfun · stepfun
step-3.5-flash
stepfun · stepfun
grok-4.20
x-ai · openrouter
Spark X2
xunfei · astroncodingplan
step-3.5-flash
stepfun · openrouter
ERNIE-4.5-Turbo
baidu · baiduqianfan
qwen3.6-plus
qwen · bailiancodingplan
doubao-seed-code
seed · volcengine-plan
Score vs Runtime
Compare model scores against average runtime. The x-axis is Avg Runtime, and the y-axis can be switched across score fields.
Pass^3
qwen3.5-plus
GLM-5.1
doubao-seed-2.0-pro
doubao-seed-2.0-code
DeepSeek-V3.2
GLM-5-Turbo
DeepSeek-V3.2
Claude Sonnet 4.6
DeepSeekV3.2
kimi-k2.5
qwen3.6-plus
mimo-v2-pro
LongCat-Flash-Thinking-2601
GLM-5
GLM-4.6
mimo-v2-omni
gpt-5.4
GLM-4.7
kat-coder-pro-v2
hunyuan-2.0-thinking
qwen3-max-2026-01-23
MiniMax-M2.5
gemini-3.1-pro-preview
hunyuan-2.0-instruct
MiniMax-M2.7
MiniMax-M2.1
Ling-2.5-1T
qwen3-coder-next
gemini-3-flash-preview
mistral-small-2603
step-3.5-flash-2603
step-3.5-flash
grok-4.20
Spark X2
step-3.5-flash
ERNIE-4.5-Turbo
qwen3.6-plus
doubao-seed-code
Avg. Seconds per Task
Score vs Cost
Compare model scores against benchmark cost. The x-axis is Cost, and the y-axis can be switched across score fields.
Pass^3
qwen3.5-plus
GLM-5.1
doubao-seed-2.0-pro
doubao-seed-2.0-code
DeepSeek-V3.2
GLM-5-Turbo
DeepSeek-V3.2
Claude Sonnet 4.6
DeepSeekV3.2
kimi-k2.5
qwen3.6-plus
mimo-v2-pro
LongCat-Flash-Thinking-2601
GLM-5
GLM-4.6
mimo-v2-omni
gpt-5.4
GLM-4.7
kat-coder-pro-v2
hunyuan-2.0-thinking
qwen3-max-2026-01-23
MiniMax-M2.5
gemini-3.1-pro-preview
hunyuan-2.0-instruct
MiniMax-M2.7
MiniMax-M2.1
Ling-2.5-1T
qwen3-coder-next
gemini-3-flash-preview
mistral-small-2603
step-3.5-flash-2603
step-3.5-flash
grok-4.20
Spark X2
step-3.5-flash
ERNIE-4.5-Turbo
qwen3.6-plus
doubao-seed-code
Cost (USD)