LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Terminal-Bench Hard (Artificial Analysis run) leaderboard

GPT-5.6 Sol leads
65.9%
models measured
126
lead over #2
+3.0 pts
median
26.5%

Leader on Terminal-Bench Hard (Artificial Analysis run) · read October 21

terminal_bench_hard_artificial_analysis_run.py


        # Terminal-Bench Hard (Artificial Analysis run), read October 21
        leaders = {
            "GPT-5.6 Sol": 65.9%,    "GPT-5.6 Terra": 62.9%,    "Claude Fable 5": 62.9%,
        }

On October 21, 2026, 126 of our catalog's models had a Terminal-Bench Hard (Artificial Analysis run) result, and OpenAI GPT-5.6 Sol led with 65.9%.

The hard subset of Terminal-Bench: tasks finished in a real terminal. Share completed. Run by Artificial Analysis on its own harness.

Role in the index

A coding board we show beside the Vibe Coding Index, not inside it: the index keeps the four boards it was defined on, so a score does not move every time a board is added. Read the methodology

126 models with a result

Original board: Artificial Analysis
Terminal-Bench Hard (Artificial Analysis run) results for models in the catalog
# Model Result Evidence Coding index Output $/Mtok
1 GPT-5.6 SolOpenAI 65.9% 76.7 $20.00
2 Claude Fable 5Anthropic 62.9% 82.1 $50.00
3 GPT-5.6 TerraOpenAI 62.9% 69.0 $12.00
4 GPT-5.5OpenAI 60.6% 67.8 $30.00
5 Claude Opus 4.8Anthropic 58.3% 66.0 $25.00
6 GPT-5.4OpenAI 57.6% 37.7 $15.00
7 Claude Opus 4.7Anthropic 54.5% 51.1 $25.00
8 Gemini 3.1 Pro PreviewGoogle 53.8% 35.3 $12.00
9 Claude Sonnet 4.6Anthropic 53.0% 57.8 $15.00
10 GPT-5.3-CodexOpenAI 53.0% 29.5 $14.00
11 GPT-5.4 MiniOpenAI 52.3% 28.2 $4.50
12 Qwen3.7 MaxQwen 50.8% 45.1 $4.425
13 GLM 5.2Z.ai 50.8% 57.7 $7.00
14 Claude Opus 4.6Anthropic 48.5% 66.9 $25.00
15 Claude Opus 4.5Anthropic 47.0% 66.8 $25.00
16 GPT-5.2OpenAI 47.0% 58.9 $14.00
17 Qwen3.7 PlusQwen 47.0% Not benchmarked $1.28
18 Gemini 3.5 FlashGoogle 46.2% 42.8 $9.00
19 GPT-5.1OpenAI 45.5% 53.3 $10.00
20 Kimi K2.7 CodeMoonshot AI 44.7% 39.1 $3.35
21 Kimi K2.6Moonshot AI 43.9% 44.2 $3.41
22 Qwen3.6 PlusQwen 43.9% 37.4 $1.95
23 MiMo-V2.5-ProXiaomi 43.2% 40.0 $0.609
24 GLM 5Z.ai 43.2% 59.7 $3.20
25 GLM 5.1Z.ai 43.2% 44.0 $4.40
26 MiniMax M3MiniMax 42.4% 57.8 $1.20
27 GPT-5.4 NanoOpenAI 42.4% Not benchmarked $1.25
28 MiMo-V2.5Xiaomi 41.7% 34.0 $0.238
29 Qwen3.5 397B A17BQwen 40.9% 28.6 $2.34
30 MiniMax M2.7MiniMax 39.4% 28.4 $1.20
31 Gemini 3 Flash PreviewGoogle 38.6% 75.8 $3.00
32 GPT-5OpenAI 37.9% 72.2 $10.00
33 Grok 4.20xAI 37.9% Not benchmarked $2.50
34 Grok 4.3xAI 37.9% 22.5 $2.50
35 GPT-5.2-CodexOpenAI 37.1% 55.1 $14.00
36 o3OpenAI 37.1% 70.7 $8.00
37 Gemma 4 31BGoogle 36.4% 38.3 $0.40
38 Nemotron 3 UltraNVIDIA 36.4% Not benchmarked $2.40
39 DeepSeek V3.2DeepSeek 35.6% 54.4 $0.42
40 Step 3.7 FlashStepFun 35.6% Not benchmarked $1.15
41 MiniMax M2.5MiniMax 34.8% 59.4 $0.95
42 Kimi K2.5Moonshot AI 34.8% 58.5 $2.50
43 GPT-5.1-CodexOpenAI 34.8% 50.5 $10.00
44 Qwen3.6 27BQwen 34.8% Not benchmarked $2.00
45 Qwen3.6 35B A3BQwen 34.8% Not benchmarked $0.95
46 Hy3 previewTencent 34.1% 22.4 $0.60
47 Mistral Medium 3.5Mistral AI 33.3% 9.2 $7.50
48 GPT-5 MiniOpenAI 33.3% 59.8 $2.00
49 GPT-5.1-Codex-MiniOpenAI 33.3% 6.2 $2.00
50 GLM 5 TurboZ.ai 33.3% Not benchmarked $4.00
51 Qwen3.5-27BQwen 32.6% 22.6 $2.16
52 Step 3.5 FlashStepFun 32.6% Not benchmarked $0.30
53 GLM 5V TurboZ.ai 32.6% 28.8 $4.00
54 DeepSeek V3.1 TerminusDeepSeek 31.8% Not benchmarked $1.00
55 GLM 4.7Z.ai 31.8% 33.6 $1.98
56 DeepSeek V3.2 ExpDeepSeek 31.1% 55.0 $0.41
57 Kimi K2 ThinkingMoonshot AI 31.1% 63.4 $2.50
58 Qwen3.5-122B-A10BQwen 31.1% 22.9 $2.08
59 MiniMax M2.1MiniMax 28.8% Not benchmarked $1.20
60 Nemotron 3 SuperNVIDIA 28.8% Not benchmarked $2.40
61 GLM 4.6Z.ai 28.8% 52.2 $2.20
62 Gemini 2.5 ProGoogle 26.5% 37.0 $10.00
63 Mercury 2Inception 26.5% 0.0 $0.75
64 Qwen3.5-35B-A3BQwen 26.5% 7.7 $1.00
65 MiniMax M2MiniMax 25.8% 45.3 $1.02
66 Command ACohere 25.0% Not benchmarked $10.00
67 Gemma 4 26B A4BGoogle 25.0% 22.8 $0.255
68 Gemini 3.1 Flash LiteGoogle 24.2% Not benchmarked $1.50
69 Qwen3.5-9BQwen 24.2% Not benchmarked $0.13
70 Kimi K2 0905Moonshot AI 23.5% Not benchmarked $2.50
71 gpt-oss-120bOpenAI 23.5% 34.5 $0.60
72 Trinity Large ThinkingArcee AI 22.7% 5.4 $0.80
73 GLM 4.5Z.ai 22.0% 64.2 $2.20
74 GLM 4.7 FlashZ.ai 22.0% Not benchmarked $0.40
75 GLM 4.5 AirZ.ai 20.5% Not benchmarked $0.85
76 Qwen3 Coder 480B A35BQwen 18.9% 50.0 $1.00
77 Qwen3 Coder NextQwen 18.2% Not benchmarked $0.80
78 Mistral Small 4Mistral AI 17.4% Not benchmarked $0.60
79 GPT-5 NanoOpenAI 17.4% 34.8 $0.40
80 R1 0528DeepSeek 15.9% 71.4 $2.15
81 Kimi K2 0711Moonshot AI 15.9% 64.7 $2.30
82 DeepSeek V3 0324DeepSeek 15.2% 42.0 $1.14
83 o4 MiniOpenAI 15.2% 68.6 $4.40
84 Qwen3 235B A22B Instruct 2507Qwen 15.2% Not benchmarked $0.55
85 Qwen3 Coder 30B A3B InstructQwen 15.2% 60.4 $0.28
86 GLM 4.6VZ.ai 14.4% Not benchmarked $0.90
87 Gemini 2.5 FlashGoogle 13.6% 28.7 $2.50
88 Nemotron 3 Nano 30B A3BNVIDIA 13.6% Not benchmarked $0.24
89 GPT-4.1OpenAI 13.6% 46.5 $8.00
90 o1OpenAI 12.9% 61.7 $60.00
91 Qwen3 VL 235B A22B InstructQwen 11.4% Not benchmarked $1.90
92 Mistral Medium 3.1Mistral AI 10.6% Not benchmarked $2.00
93 gpt-oss-20bOpenAI 10.6% Not benchmarked $0.09
94 Qwen3 Next 80B A3B InstructQwen 9.8% Not benchmarked $1.10
95 GPT-4oOpenAI 8.3% 38.8 $10.00
96 GPT-4o (2024-08-06)OpenAI 8.3% 23.1 $10.00
97 GPT-4.1 MiniOpenAI 7.6% 28.5 $1.60
98 Solar Pro 3Upstage 7.6% Not benchmarked $0.60
99 Nova Premier 1.0Amazon 6.8% Not benchmarked $12.50
100 Llama 4 MaverickMeta Llama 6.8% 18.1 $0.6525
101 o3 MiniOpenAI 6.8% 52.1 $4.40
102 Qwen3 30B A3BQwen 6.8% Not benchmarked $0.50
103 GLM 4.5VZ.ai 6.8% Not benchmarked $1.80
104 Nova Pro 1.0Amazon 6.1% Not benchmarked $3.20
105 Qwen3 30B A3B Instruct 2507Qwen 6.1% Not benchmarked $0.30
106 Qwen3 VL 30B A3B InstructQwen 6.1% Not benchmarked $0.60
107 Qwen3 14BQwen 5.3% Not benchmarked $0.24
108 Gemini 2.5 Flash LiteGoogle 4.5% Not benchmarked $0.40
109 Qwen2.5 72B InstructQwen 4.5% Not benchmarked $0.40
110 Gemma 3 27BGoogle 3.8% 4.9 $0.35
111 Phi 4Microsoft 3.8% Not benchmarked $0.14
112 Mistral Medium 3Mistral AI 3.8% Not benchmarked $2.00
113 GPT-4.1 NanoOpenAI 3.8% 8.9 $0.40
114 Qwen3 VL 8B InstructQwen 3.8% Not benchmarked $0.75
115 Llama 3.1 70B InstructMeta Llama 3.0% Not benchmarked $0.40
116 Llama 3.3 70B InstructMeta Llama 3.0% Not benchmarked $0.32
117 Qwen3 32BQwen 3.0% 40.0 $0.28
118 Nova Micro 1.0Amazon 1.5% Not benchmarked $0.14
119 Granite 4.0 MicroIBM Granite 1.5% Not benchmarked $0.112
120 Llama 4 ScoutMeta Llama 1.5% 9.1 $0.30
121 Nova Lite 1.0Amazon 0.8% Not benchmarked $0.24
122 Gemma 3 12BGoogle 0.8% Not benchmarked $0.15
123 Gemma 3 4BGoogle 0.8% Not benchmarked $0.05
124 Llama 3.1 8B InstructMeta Llama 0.8% Not benchmarked $0.08
125 Llama 3.2 1B InstructMeta Llama 0.0% Not benchmarked $0.201
126 Reka Flash 3Reka 0.0% Not benchmarked $0.20

Only models in our catalog are listed, and each appears once with its best published run. 195 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.