cyberclaw.directory
Sign in

OpenClaw Rankings

Best models for OpenClaw usage

OpenClaw Model Rankings

Top 10 of 87 qualifying models

Score weights:45% Terminal-Bench40% Capability15% Cost
#1GPT-5.4 (xhigh)

OpenAI

73.6

OpenClaw

Intelligence · Coding

Int 51.4·Code 71.1

Terminal-Bench Hard

57.6%

Latency (TTFT)

97.04 s

Run Cost

$0.090

in $2.500 / out $15.000 per 1M

#2GLM-5.2 (max)

Z AI

72.6

OpenClaw

Intelligence · Coding

Int 51.1·Code 68.8

Terminal-Bench Hard

50.8%

Latency (TTFT)

856 ms

Run Cost

$0.034

in $1.400 / out $4.400 per 1M

#3Gemini 3.1 Pro Preview

Google

70.2

OpenClaw

Intelligence · Coding

Int 46.5·Code 68.8

Terminal-Bench Hard

53.8%

Latency (TTFT)

20.86 s

Run Cost

$0.072

in $2.000 / out $12.000 per 1M

#4GPT-5.4 mini (xhigh)

OpenAI

70.0

OpenClaw

Intelligence · Coding

Int 40.0·Code 56.1

Terminal-Bench Hard

52.3%

Latency (TTFT)

10.63 s

Run Cost

$0.027

in $0.750 / out $4.500 per 1M

#5Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)

Anthropic

69.8

OpenClaw

Intelligence · Coding

Int 59.9·Code 76.5

Terminal-Bench Hard

62.9%

Latency (TTFT)

46.22 s

Run Cost

$0.320

in $10.000 / out $50.000 per 1M

#6Claude Opus 4.8 (Adaptive Reasoning, Max Effort)

Anthropic

69.6

OpenClaw

Intelligence · Coding

Int 55.7·Code 74.3

Terminal-Bench Hard

58.3%

Latency (TTFT)

19.19 s

Run Cost

$0.160

in $5.000 / out $25.000 per 1M

#7Qwen3.7 Max

Alibaba

68.4

OpenClaw

Intelligence · Coding

Int 46.0·Code 66.0

Terminal-Bench Hard

50.8%

Latency (TTFT)

1.59 s

Run Cost

$0.060

in $2.500 / out $7.500 per 1M

#8DeepSeek V4 Pro (Reasoning, Max Effort)

DeepSeek

68.4

OpenClaw

Intelligence · Coding

Int 44.3·Code 59.4

Terminal-Bench Hard

46.2%

Latency (TTFT)

1.16 s

Run Cost

$0.0087

in $0.435 / out $0.870 per 1M

#9GPT-5.5 (xhigh)

OpenAI

67.9

OpenClaw

Intelligence · Coding

Int 54.8·Code 74.9

Terminal-Bench Hard

60.6%

Latency (TTFT)

32.93 s

Run Cost

$0.180

in $5.000 / out $30.000 per 1M

#10Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Anthropic

67.6

OpenClaw

Intelligence · Coding

Int 47.2·Code 63.0

Terminal-Bench Hard

53.0%

Latency (TTFT)

41.22 s

Run Cost

$0.096

in $3.000 / out $15.000 per 1M

Scores (0–100) are percentile-normalized across all qualifying models — not raw benchmark percentages. Standard run = 12,000 input + 4,000 output tokens. Hover column headers for metric definitions. Data via Artificial Analysis.

Value Map

Best models cluster top-left: high Intelligence Index, low run cost. Bubble size scales with Terminal-Bench Hard score. Hover a bubble or legend entry to inspect.

Scroll horizontally to explore →

← cheaper · smarter ↑ (ideal zone)$0.0010$0.0041$0.018$0.075$0.320521375369Run Cost per call, log scale →Intelligence Index →GPT-5.4 (xhigh) Provider: OpenAI OpenClaw Score: 73.6 Intelligence Index: 51.4 Agent Fitness: 53.4 Terminal-Bench Hard: 57.6% Latency (TTFT): 97.04 s Run cost: $0.090GLM-5.2 (max) Provider: Z AI OpenClaw Score: 72.6 Intelligence Index: 51.1 Agent Fitness: 52.9 Terminal-Bench Hard: 50.8% Latency (TTFT): 856 ms Run cost: $0.034Gemini 3.1 Pro Preview Provider: Google OpenClaw Score: 70.2 Intelligence Index: 46.5 Agent Fitness: 48.7 Terminal-Bench Hard: 53.8% Latency (TTFT): 20.86 s Run cost: $0.072GPT-5.4 mini (xhigh) Provider: OpenAI OpenClaw Score: 70.0 Intelligence Index: 40.0 Agent Fitness: 41.6 Terminal-Bench Hard: 52.3% Latency (TTFT): 10.63 s Run cost: $0.027Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) Provider: Anthropic OpenClaw Score: 69.8 Intelligence Index: 59.9 Agent Fitness: 61.6 Terminal-Bench Hard: 62.9% Latency (TTFT): 46.22 s Run cost: $0.320Claude Opus 4.8 (Adaptive Reasoning, Max Effort) Provider: Anthropic OpenClaw Score: 69.6 Intelligence Index: 55.7 Agent Fitness: 57.6 Terminal-Bench Hard: 58.3% Latency (TTFT): 19.19 s Run cost: $0.160Qwen3.7 Max Provider: Alibaba OpenClaw Score: 68.4 Intelligence Index: 46.0 Agent Fitness: 48.0 Terminal-Bench Hard: 50.8% Latency (TTFT): 1.59 s Run cost: $0.060DeepSeek V4 Pro (Reasoning, Max Effort) Provider: DeepSeek OpenClaw Score: 68.4 Intelligence Index: 44.3 Agent Fitness: 45.8 Terminal-Bench Hard: 46.2% Latency (TTFT): 1.16 s Run cost: $0.0087GPT-5.5 (xhigh) Provider: OpenAI OpenClaw Score: 67.9 Intelligence Index: 54.8 Agent Fitness: 56.8 Terminal-Bench Hard: 60.6% Latency (TTFT): 32.93 s Run cost: $0.180Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) Provider: Anthropic OpenClaw Score: 67.6 Intelligence Index: 47.2 Agent Fitness: 48.8 Terminal-Bench Hard: 53.0% Latency (TTFT): 41.22 s Run cost: $0.096GPT-5.5 (high) Provider: OpenAI OpenClaw Score: 67.1 Intelligence Index: 53.1 Agent Fitness: 55.0 Terminal-Bench Hard: 59.8% Latency (TTFT): 16.52 s Run cost: $0.180Qwen3.7 Plus Provider: Alibaba OpenClaw Score: 66.9 Intelligence Index: 39.0 Agent Fitness: 40.7 Terminal-Bench Hard: 47.0% Latency (TTFT): 1.91 s Run cost: $0.0094GPT-5.5 (medium) Provider: OpenAI OpenClaw Score: 66.2 Intelligence Index: 50.4 Agent Fitness: 52.5 Terminal-Bench Hard: 57.6% Latency (TTFT): 5.29 s Run cost: $0.180MiniMax-M3 Provider: MiniMax OpenClaw Score: 65.3 Intelligence Index: 44.4 Agent Fitness: 45.8 Terminal-Bench Hard: 42.4% Latency (TTFT): 1.59 s Run cost: $0.0084MiMo-V2.5-Pro Provider: Xiaomi OpenClaw Score: 65.2 Intelligence Index: 42.2 Agent Fitness: 44.0 Terminal-Bench Hard: 43.2% Latency (TTFT): 2.01 s Run cost: $0.0087Kimi K2.6 Provider: Kimi OpenClaw Score: 64.9 Intelligence Index: 44.2 Agent Fitness: 46.0 Terminal-Bench Hard: 43.9% Latency (TTFT): 1.14 s Run cost: $0.027Kimi K2.7 Code Provider: Kimi OpenClaw Score: 64.7 Intelligence Index: 41.9 Agent Fitness: 43.8 Terminal-Bench Hard: 44.7% Latency (TTFT): 1.29 s Run cost: $0.027Claude Opus 4.7 (Adaptive Reasoning, Max Effort) Provider: Anthropic OpenClaw Score: 63.8 Intelligence Index: 53.5 Agent Fitness: 55.5 Terminal-Bench Hard: 51.5% Latency (TTFT): 17.58 s Run cost: $0.160Qwen3.6 Plus Provider: Alibaba OpenClaw Score: 63.8 Intelligence Index: 39.6 Agent Fitness: 41.1 Terminal-Bench Hard: 43.9% Latency (TTFT): 1.66 s Run cost: $0.018GPT-5.4 nano (xhigh) Provider: OpenAI OpenClaw Score: 63.1 Intelligence Index: 38.2 Agent Fitness: 40.0 Terminal-Bench Hard: 42.4% Latency (TTFT): 2.77 s Run cost: $0.0074Gemini 3.5 Flash (high) Provider: Google OpenClaw Score: 62.6 Intelligence Index: 50.2 Agent Fitness: 52.2 Terminal-Bench Hard: 40.9% Latency (TTFT): 12.75 s Run cost: $0.054GLM-5.1 (Reasoning) Provider: Z AI OpenClaw Score: 62.0 Intelligence Index: 40.2 Agent Fitness: 41.8 Terminal-Bench Hard: 43.2% Latency (TTFT): 773 ms Run cost: $0.034GPT-5.1 (high) Provider: OpenAI OpenClaw Score: 60.6 Intelligence Index: 36.9 Agent Fitness: 38.1 Terminal-Bench Hard: 45.5% Latency (TTFT): 26.96 s Run cost: $0.055MiniMax-M2.7 Provider: MiniMax OpenClaw Score: 60.4 Intelligence Index: 38.1 Agent Fitness: 39.6 Terminal-Bench Hard: 39.4% Latency (TTFT): 1.32 s Run cost: $0.0084GPT-5.5 (low) Provider: OpenAI OpenClaw Score: 58.9 Intelligence Index: 43.5 Agent Fitness: 45.2 Terminal-Bench Hard: 52.3% Latency (TTFT): 1.42 s Run cost: $0.180DeepSeek V4 Flash (Reasoning, Max Effort) Provider: DeepSeek OpenClaw Score: 58.7 Intelligence Index: 40.3 Agent Fitness: 41.9 Terminal-Bench Hard: 35.6% Latency (TTFT): 998 ms Run cost: $0.0028Qwen3.5 397B A17B (Reasoning) Provider: Alibaba OpenClaw Score: 58.6 Intelligence Index: 33.7 Agent Fitness: 35.2 Terminal-Bench Hard: 40.9% Latency (TTFT): 1.79 s Run cost: $0.022Grok 4.3 (high) Provider: xAI OpenClaw Score: 57.0 Intelligence Index: 37.6 Agent Fitness: 38.1 Terminal-Bench Hard: 37.9% Latency (TTFT): 23.25 s Run cost: $0.025Nemotron 3 Ultra 550B A55B (Reasoning) Provider: NVIDIA OpenClaw Score: 56.7 Intelligence Index: 37.8 Agent Fitness: 38.9 Terminal-Bench Hard: 36.4% Latency (TTFT): 869 ms Run cost: $0.019Qwen3.6 27B (Reasoning) Provider: Alibaba OpenClaw Score: 55.2 Intelligence Index: 37.1 Agent Fitness: 38.8 Terminal-Bench Hard: 34.8% Latency (TTFT): 1.51 s Run cost: $0.022Qwen3.6 35B A3B (Reasoning) Provider: Alibaba OpenClaw Score: 53.9 Intelligence Index: 31.6 Agent Fitness: 32.6 Terminal-Bench Hard: 34.8% Latency (TTFT): 1.36 s Run cost: $0.0089Step 3.7 Flash Provider: StepFun OpenClaw Score: 53.8 Intelligence Index: 29.7 Agent Fitness: 30.5 Terminal-Bench Hard: 35.6% Latency (TTFT): 747 ms Run cost: $0.0070GPT-5.5 (Non-reasoning) Provider: OpenAI OpenClaw Score: 53.4 Intelligence Index: 35.4 Agent Fitness: 37.5 Terminal-Bench Hard: 49.2% Latency (TTFT): 881 ms Run cost: $0.180GLM-4.7 (Reasoning) Provider: Z AI OpenClaw Score: 51.6 Intelligence Index: 33.7 Agent Fitness: 34.9 Terminal-Bench Hard: 31.8% Latency (TTFT): 550 ms Run cost: $0.016Qwen3.5 122B A10B (Reasoning) Provider: Alibaba OpenClaw Score: 50.4 Intelligence Index: 32.3 Agent Fitness: 33.6 Terminal-Bench Hard: 31.1% Latency (TTFT): 1.26 s Run cost: $0.018GPT-5 mini (high) Provider: OpenAI OpenClaw Score: 49.1 Intelligence Index: 25.3 Agent Fitness: 24.3 Terminal-Bench Hard: 33.3% Latency (TTFT): 95.48 s Run cost: $0.011Claude 4.5 Sonnet (Reasoning) Provider: Anthropic OpenClaw Score: 49.0 Intelligence Index: 36.4 Agent Fitness: 38.0 Terminal-Bench Hard: 35.6% Latency (TTFT): 8.44 s Run cost: $0.096GPT-5 (high) Provider: OpenAI OpenClaw Score: 48.9 Intelligence Index: 34.7 Agent Fitness: 35.0 Terminal-Bench Hard: 32.6% Latency (TTFT): 80.59 s Run cost: $0.055Mistral Medium 3.5 Provider: Mistral OpenClaw Score: 48.7 Intelligence Index: 29.9 Agent Fitness: 31.6 Terminal-Bench Hard: 33.3% Latency (TTFT): 644 ms Run cost: $0.048Ring-2.6-1T Provider: InclusionAI OpenClaw Score: 48.2 Intelligence Index: 30.6 Agent Fitness: 31.8 Terminal-Bench Hard: 28.8% Latency (TTFT): 1.82 s Run cost: $0.014NVIDIA Nemotron 3 Super 120B A12B (Reasoning) Provider: NVIDIA OpenClaw Score: 46.8 Intelligence Index: 25.4 Agent Fitness: 26.6 Terminal-Bench Hard: 28.8% Latency (TTFT): 980 ms Run cost: $0.0061Claude 4.5 Haiku (Reasoning) Provider: Anthropic OpenClaw Score: 45.0 Intelligence Index: 29.6 Agent Fitness: 31.0 Terminal-Bench Hard: 27.3% Latency (TTFT): 10.46 s Run cost: $0.032GLM-4.6 (Reasoning) Provider: Z AI OpenClaw Score: 44.4 Intelligence Index: 28.7 Agent Fitness: 30.4 Terminal-Bench Hard: 25.0% Latency (TTFT): 998 ms Run cost: $0.015Gemini 3.1 Flash-Lite Provider: Google OpenClaw Score: 42.5 Intelligence Index: 25.0 Agent Fitness: 26.0 Terminal-Bench Hard: 24.2% Latency (TTFT): 4.47 s Run cost: $0.0090gpt-oss-120b (high) Provider: OpenAI OpenClaw Score: 41.8 Intelligence Index: 23.8 Agent Fitness: 24.5 Terminal-Bench Hard: 23.5% Latency (TTFT): 497 ms Run cost: $0.0042Qwen3.5 9B (Reasoning) Provider: Alibaba OpenClaw Score: 41.7 Intelligence Index: 21.4 Agent Fitness: 22.1 Terminal-Bench Hard: 24.2% Latency (TTFT): 1.04 s Run cost: $0.0018Qwen3.6 27B (Non-reasoning) Provider: Alibaba OpenClaw Score: 41.4 Intelligence Index: 30.5 Agent Fitness: 32.1 Terminal-Bench Hard: 21.2% Latency (TTFT): 1.47 s Run cost: $0.022Gemini 2.5 Pro Provider: Google OpenClaw Score: 40.6 Intelligence Index: 25.8 Agent Fitness: 26.6 Terminal-Bench Hard: 26.5% Latency (TTFT): 18.24 s Run cost: $0.055HyperNova 60B 2605 Provider: Multiverse Computing OpenClaw Score: 39.6 Intelligence Index: 17.8 Agent Fitness: 18.3 Terminal-Bench Hard: 23.5% Latency (TTFT): 740 ms Run cost: $0.0010Nova 2.0 Pro Preview (medium) Provider: Amazon OpenClaw Score: 37.3 Intelligence Index: 21.8 Agent Fitness: 23.0 Terminal-Bench Hard: 24.2% Latency (TTFT): 14.96 s Run cost: $0.055Grok 4.3 (Non-reasoning) Provider: xAI OpenClaw Score: 36.7 Intelligence Index: 24.8 Agent Fitness: 25.8 Terminal-Bench Hard: 18.9% Latency (TTFT): 555 ms Run cost: $0.025Ling 2.6 Flash Provider: InclusionAI OpenClaw Score: 36.4 Intelligence Index: 14.1 Agent Fitness: 15.2 Terminal-Bench Hard: 21.2% Latency (TTFT): 917 ms Run cost: $0.0024Qwen3 Coder Next Provider: Alibaba OpenClaw Score: 36.3 Intelligence Index: 21.1 Agent Fitness: 22.6 Terminal-Bench Hard: 18.2% Latency (TTFT): 1.01 s Run cost: $0.0090Qwen3.5 4B (Reasoning) Provider: Alibaba OpenClaw Score: 36.1 Intelligence Index: 20.1 Agent Fitness: 20.4 Terminal-Bench Hard: 18.2% Latency (TTFT): 605 ms Run cost: $0.0010Mistral Small 4 (Reasoning) Provider: Mistral OpenClaw Score: 35.2 Intelligence Index: 19.6 Agent Fitness: 20.3 Terminal-Bench Hard: 17.4% Latency (TTFT): 546 ms Run cost: $0.0042KAT-Coder-Pro V1 Provider: KwaiKAT OpenClaw Score: 34.7 Intelligence Index: 34.6 Agent Fitness: 37.0 Terminal-Bench Hard: 9.1% Latency (TTFT): 1.17 s Run cost: $0.0084Nova 2.0 Lite (high) Provider: Amazon OpenClaw Score: 33.1 Intelligence Index: 18.2 Agent Fitness: 18.7 Terminal-Bench Hard: 16.7% Latency (TTFT): 19.76 s Run cost: $0.014Mistral Large 3 Provider: Mistral OpenClaw Score: 31.6 Intelligence Index: 15.9 Agent Fitness: 16.3 Terminal-Bench Hard: 15.9% Latency (TTFT): 655 ms Run cost: $0.012Nova 2.0 Pro Preview (low) Provider: Amazon OpenClaw Score: 30.6 Intelligence Index: 19.6 Agent Fitness: 20.2 Terminal-Bench Hard: 17.4% Latency (TTFT): 4.30 s Run cost: $0.055NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) Provider: NVIDIA OpenClaw Score: 29.9 Intelligence Index: 14.2 Agent Fitness: 14.2 Terminal-Bench Hard: 13.6% Latency (TTFT): 1.39 s Run cost: $0.0014

Data provided by Artificial Analysis