All General Coding Reasoning Multimodal Safety Specialized Agentic Language Long Context Speed Text-to-Image Image Editing Text-to-Video Image-to-Video
BFCL83 BrowseComp OSWorld-Verified TAU-bench18 TAU-bench Airline22 TAU-bench Banking Knowledge27 TAU-bench Retail23 TAU-bench Telecom18 Tau Banking (Artificial Analysis)209 Tau2 Telecom (Artificial Analysis)439 Terminal-Bench 2.1241 Terminal-Bench Hard431 ToolBench WebArena

Terminal-Bench 2.1

Agentic

Source ↗ Higher is better

Rank Model Provider Score
1 Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) Anthropic
91.4%
2 Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback) Anthropic
91.0%
3 GPT-6 Astra (high) OpenAI
89.9%
4 Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback) Anthropic
89.9%
5 GPT-5.6 Sol (xhigh) OpenAI
89.5%
6 GPT-6 Astra (medium) OpenAI
89.5%
7 GPT-6 Astra (xhigh) OpenAI
89.1%
8 Claude Opus 5 (Adaptive Reasoning, Max Effort) Anthropic
89.1%
9 Qwen: Qwen3.8 Max (0902) Alibaba
88.8%
10 Grok 4.6 (high) SpaceXAI
88.4%
11 GPT-6 Astra (max) OpenAI
88.4%
12 Grok 4.6 (xhigh) SpaceXAI
88.0%
13 GPT-5.6 Sol (max) OpenAI
88.0%
14 GPT-5.6 Terra (max) OpenAI
88.0%
15 Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) Anthropic
88.0%
16 GPT-6 Astra (low) OpenAI
88.0%
17 Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) Anthropic
88.0%
18 Claude Opus 5 (Adaptive Reasoning, High Effort) Anthropic
87.6%
19 Gemini 3.8 Flash (high) Google
87.6%
20 GPT-5.6 Sol (high) OpenAI
87.3%
21 Claude Opus 5 (Adaptive Reasoning, Medium Effort) Anthropic
86.1%
22 GPT-5.6 Sol (medium) OpenAI
86.1%
23 Qwen3.8-Flash-Next Alibaba
86.1%
24 Gemini 3.7 Flash (high) Google
85.8%
25 Muse Spark 1.3 (xhigh) Meta
85.4%
26 Kimi K3 (max) Moonshot AI
85.0%
27 Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback) Anthropic
85.0%
28 Claude Opus 4.8 (Adaptive Reasoning, Max Effort) Anthropic
84.6%
29 Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) Anthropic
84.6%
30 Muse Spark 1.3 (max) Meta
84.3%
31 Z.ai: GLM 5.3 Flash Z.ai
84.3%
32 GPT-5.5 (xhigh) OpenAI
84.3%
33 Grok 4.6 (medium) SpaceXAI
84.3%
34 GLM-5.3 (max) Z.ai
83.9%
35 Gemini 3.8 Flash (medium) Google
83.9%
36 Gemini 3.8 Flash (low) Google
83.1%
37 Claude Opus 4.7 (Adaptive Reasoning, Max Effort) Anthropic
83.1%
38 Kimi K3 (low) Moonshot AI
82.4%
39 Qwen: Qwen3.8 2.4T A95B Alibaba
82.0%
40 Grok 4.5 (high) SpaceXAI
81.6%
41 Qwen3.8 Max Alibaba
81.3%
42 GPT-5.6 Luna (max) OpenAI
80.9%
43 GPT-5.5 (medium) OpenAI
80.5%
44 Claude Sonnet 5 (Adaptive Reasoning, Max Effort) Anthropic
80.5%
45 GPT-5.6 Terra (xhigh) OpenAI
80.1%
46 Muse Spark 1.2 (xhigh) Meta
80.1%
47 Gemini 3.7 Flash (low) Google
79.8%
48 Qwen3.8 27B (xhigh) Alibaba
79.8%
49 GPT-5.5 (high) OpenAI
79.4%
50 DeepSeek V4 Pro 0813 (Reasoning, Max Effort) DeepSeek
78.7%

Access this data

JSON GET /api/v1 
[
  {
    "rank": 1,
    "model": "Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)",
    "provider": "Anthropic",
    "score": "91.3857677902622"
  },
  {
    "rank": 2,
    "model": "Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)",
    "provider": "Anthropic",
    "score": "91.0112359550562"
  },
  {
    "rank": 3,
    "model": "GPT-6 Astra (high)",
    "provider": "OpenAI",
    "score": "89.8876404494382"
  }
]
Curl 
$ curl -s https://rightrank.ai/api/v1/rankings/terminal-bench-2-1.json | jq '.[:3]'
CLI brew install rightrank 
$ rightrank top terminal-bench-2-1 --limit 5
$ rightrank compare claude-opus-4.6 gpt-5.4-pro --category terminal-bench-2-1
$ rightrank value terminal-bench-2-1  # best performance per dollar
Claude Code Skill Skill 
# In Claude Code, use the RightRank skill:
/rightrank best model for terminal-bench-2-1
/rightrank compare top 5 by value
/rightrank cheapest model scoring above 80 on terminal-bench-2-1