Benchmark Almanac

Terminal-Bench 2.0

GPT-5.5 84.7% leads since April 2026
GPT-5.5 84.7% leads since April 2026
In use

89 hard terminal tasks across 16 categories in containers. Hidden tests verify the final state. 5 trials per task.

Leaderboard entries are audited and trajectories are public. Later versions exist (Vals now uses Terminal-Bench 4.0).

Released
November 2025
Kind
Terminal tasks
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Strongest model tested
GPT-5.5 #20 on the capability index
Last in a launch post
July 2026
Last model added
July 2026
Scores refreshed
November 12th, 2026

In the authors’ words

AI agents may soon become capable of autonomously completing valuable, long-horizon tasks in diverse domains. Current benchmarks either do not measure real-world tasks, or are not sufficiently difficult to meaningfully measure frontier models. To this end, we present Terminal-Bench 2.0: a carefully curated hard benchmark composed of 89 tasks in computer terminal environments inspired by problems from real workflows. Each task features a unique environment, human-written solution, and comprehensive tests for verification. We show that frontier models and agents score less than 65% on the benchmark and conduct an error analysis to identify areas for model and agent improvement. We publish the dataset and evaluation harness to assist developers and researchers in future work at https://www.tbench.ai/ .

Abstract of Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces (arXiv:2601.11868, CC0)
46
Entries on its own leaderboard, 43 models we track
62
Launch posts list it, from 21 labs
28
System cards report it, from 17 labs

GPT-5.5 leads with 84.7% among models released by April 2026

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

GPT-5.5 leads with 84.7%, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

46 entries. Scroll sideways for the rest.

  1. 84.7 GPT-5.5
  2. 80.2 Gemini 3.1 Pro
  3. 80.2 Claude Opus 4.7
  4. 78.4 GPT-5.3 Codex
  5. 76.4 Claude Opus 4.6
  6. 69.4 Gemini 3 Pro
  7. 66.5 GPT-5.2 Codex
  8. 64.9 GPT-5.2
  9. 64.3 Gemini 3 Flash
  10. 63.1 Claude Opus 4.5
  11. 61.6 GPT-5.1-Codex-mini
  12. 60.5 GPT-5.1-Codex-Max
  13. 57.8 GPT-5.1-Codex
  14. 57.3 Grok 4.20
  15. 53.4 Claude Sonnet 4.6
  16. 52.4 GLM-5
  17. 49.6 GPT-5
  18. 47.6 GPT-5.1
  19. 46.5 Claude Sonnet 4.5
  20. 45.1 MiniMax-M2.7
  21. 44.3 GPT-5-Codex
  22. 43.2 Kimi K2.5
  23. 42.9 MiniMax M2.7 Highspeed
  24. 42.7 MiniMax-M2.5
  25. 39.5 DeepSeek-V3.2
  26. 38.0 Claude Opus 4.1
  27. 36.6 MiniMax-M2.1
  28. 35.7 Kimi K2 Thinking
  29. 35.5 Claude Haiku 4.5
  30. 34.8 GPT-5 mini
  31. 33.4 GLM-4.7
  32. 32.6 Gemini 2.5 Pro
  33. 30.0 MiniMax-M2
  34. 27.8 Kimi K2
  35. 27.2 Grok 4
  36. 27.2 Qwen3-Coder-480B-A35B
  37. 25.8 grok-code-fast-1
  38. 24.6 Qwen 3.6 35B-A3B
  39. 24.5 GLM-4.6
  40. 21.8 GPT-5 nano
  41. 19.3 TermiGen-32B
  42. 18.7 gpt-oss-120b
  43. 17.1 Gemini 2.5 Flash
  44. 17.0 AfterQuery-GPT-OSS-20B
  45. 9.21 Qwen3.5-9B
  46. 3.37 gpt-oss-20b
  1. 1 GPT-5.5 84.7%
  2. 2 Gemini 3.1 Pro 80.2%
  3. 3 Claude Opus 4.7 80.2%
  4. 4 GPT-5.3 Codex 78.4%
  5. 5 Claude Opus 4.6 76.4%
  6. 6 Gemini 3 Pro 69.4%
  7. 7 GPT-5.2 Codex 66.5%
  8. 8 GPT-5.2 64.9%
  9. 9 Gemini 3 Flash 64.3%
  10. 10 Claude Opus 4.5 63.1%
  11. 11 GPT-5.1-Codex-mini 61.6%
  12. 12 GPT-5.1-Codex-Max 60.5%
  13. 13 GPT-5.1-Codex 57.8%
  14. 14 Grok 4.20 reasoning 57.3%
  15. 15 Claude Sonnet 4.6 53.4%
  16. 16 GLM-5 52.4%
  17. 17 GPT-5 49.6%
  18. 18 GPT-5.1 47.6%
  19. 19 Claude Sonnet 4.5 46.5%
  20. 20 MiniMax-M2.7 45.1%
  21. 21 GPT-5-Codex 44.3%
  22. 22 Kimi K2.5 43.2%
  23. 23 MiniMax M2.7 Highspeed 42.9%
  24. 24 MiniMax-M2.5 42.7%
  25. 25 DeepSeek-V3.2 39.5%
  26. 26 Claude Opus 4.1 38%
  27. 27 MiniMax-M2.1 36.6%
  28. 28 Kimi K2 Thinking 35.7%
  29. 29 Claude Haiku 4.5 35.5%
  30. 30 GPT-5 mini 34.8%
  31. 31 GLM-4.7 33.4%
  32. 32 Gemini 2.5 Pro Jun 2025 32.6%
  33. 33 MiniMax-M2 30%
  34. 34 Kimi K2 Sep 2025 27.8%
  35. 35 Grok 4 27.2%
  36. 36 Qwen3-Coder-480B-A35B 27.2%
  37. 37 grok-code-fast-1 25.8%
  38. 38 Qwen 3.6 35B-A3B 24.6%
  39. 39 GLM-4.6 24.5%
  40. 40 GPT-5 nano 21.8%
  41. 41 TermiGen-32B 19.3%
  42. 42 gpt-oss-120b 18.7%
  43. 43 Gemini 2.5 Flash Jun 2025 17.1%
  44. 44 AfterQuery-GPT-OSS-20B 17%
  45. 45 Qwen3.5-9B 9.21%
  46. 46 gpt-oss-20b 3.37%

Source: Official Terminal-Bench 2.0 leaderboard

87 models. Scroll sideways for the rest.

  1. 84.7 GPT-5.5
  2. 82.0 Claude Mythos Preview
  3. 80.2 Gemini 3.1 Pro
  4. 80.2 Claude Opus 4.7
  5. 78.4 GPT-5.3 Codex
  6. 76.4 Claude Opus 4.6
  7. 70.3 Qwen3.7-Plus
  8. 69.7 Qwen3.7-Max
  9. 69.4 Gemini 3 Pro
  10. 69.3 Composer 2.5
  11. 69.0 GLM-5.1
  12. 68.4 MiMo-V2.5-Pro
  13. 67.9 DeepSeek-V4-Pro
  14. 66.7 Kimi K2.6
  15. 66.5 GPT-5.2 Codex
  16. 65.8 MiMo-V2.5
  17. 65.4 Qwen 3.6 Max (Preview)
  18. 64.9 GPT-5.2
  19. 64.3 Gemini 3 Flash
  20. 63.1 Claude Opus 4.5
  21. 61.7 Composer 2
  22. 61.6 Qwen 3.6 Plus
  23. 61.6 GPT-5.1-Codex-mini
  24. 60.5 GPT-5.1-Codex-Max
  25. 60.0 GPT-5.4 Mini
  26. 59.3 Qwen3.6 27B
  27. 59.1 Claude Sonnet 4.6
  28. 59.0 Muse Spark
  29. 58.4 GPT-5.3-Codex-Spark
  30. 57.8 GPT-5.1-Codex
  31. 57.3 Grok 4.20
  32. 57.1 MiMo-V2-Pro
  33. 57.0 MiniMax-M2.7
  34. 56.9 DeepSeek-V4-Flash
  35. 56.2 GLM-5
  36. 55.8 Seed2.0 Pro
  37. 54.8 MAI-Code-1-Flash
  38. 54.4 Tencent Hy3 preview
  39. 52.5 Qwen3.5 397B-A17B
  40. 51.7 MiniMax-M2.5
  41. 51.5 Qwen 3.6 35B-A3B
  42. 51.0 Step 3.5 Flash
  43. 50.8 Kimi K2.5
  44. 49.6 GPT-5
  45. 49.4 Qwen3.5-122B-A10B
  46. 47.9 MiniMax-M2.1
  47. 47.9 Composer 1.5
  48. 47.6 GPT-5.1
  49. 46.5 Claude Sonnet 4.5
  50. 46.4 DeepSeek-V3.2
  51. 46.3 GPT-5.4 Nano
  52. 46.0 MAI-Thinking-1
  53. 45.8 Laguna M.1
  54. 45.2 Seed1.8
  55. 44.3 GPT-5-Codex
  56. 43.3 Seed2.0 Lite (0428)
  57. 41.6 Qwen3.5-27B
  58. 41.0 GLM-4.7
  59. 40.5 Qwen3.5-35B-A3B
  60. 40.5 Ling-2.6-1T
  61. 38.5 MiMo-V2-Flash
  62. 38.0 Claude Opus 4.1
  63. 37.5 Laguna XS 2.1
  64. 36.2 Qwen3-Coder-Next
  65. 36.0 North Mini Code 1.0
  66. 35.7 Kimi K2 Thinking
  67. 35.7 Laguna XS.2
  68. 35.5 Claude Haiku 4.5
  69. 34.8 GPT-5 mini
  70. 32.6 Gemini 2.5 Pro
  71. 32.6 Devstral 2
  72. 31.0 Nemotron 3 Super
  73. 30.0 MiniMax-M2
  74. 29.0 K-EXAONE
  75. 27.8 Kimi K2
  76. 27.2 Grok 4
  77. 27.2 Qwen3-Coder-480B-A35B
  78. 25.8 grok-code-fast-1
  79. 24.5 GLM-4.6
  80. 22.5 Devstral Small 2
  81. 21.8 GPT-5 nano
  82. 21.1 Nemotron-Cascade 2 30B-A3B
  83. 18.7 gpt-oss-120b
  84. 17.1 Gemini 2.5 Flash
  85. 10.1 Solar Pro 3 (March 2026 update)
  86. 9.21 Qwen3.5-9B
  87. 3.37 gpt-oss-20b
  1. 1 GPT-5.5 84.7%
  2. 2 Claude Mythos Preview 82%
  3. 3 Gemini 3.1 Pro 80.2%
  4. 4 Claude Opus 4.7 80.2%
  5. 5 GPT-5.3 Codex 78.4%
  6. 6 Claude Opus 4.6 76.4%
  7. 7 Qwen3.7-Plus 70.3%
  8. 8 Qwen3.7-Max 69.7%
  9. 9 Gemini 3 Pro 69.4%
  10. 10 Composer 2.5 69.3%
  11. 11 GLM-5.1 69%
  12. 12 MiMo-V2.5-Pro 68.4%
  13. 13 DeepSeek-V4-Pro 67.9%
  14. 14 Kimi K2.6 66.7%
  15. 15 GPT-5.2 Codex 66.5%
  16. 16 MiMo-V2.5 65.8%
  17. 17 Qwen 3.6 Max (Preview) 65.4%
  18. 18 GPT-5.2 64.9%
  19. 19 Gemini 3 Flash 64.3%
  20. 20 Claude Opus 4.5 63.1%
  21. 21 Composer 2 61.7%
  22. 22 Qwen 3.6 Plus 61.6%
  23. 23 GPT-5.1-Codex-mini 61.6%
  24. 24 GPT-5.1-Codex-Max 60.5%
  25. 25 GPT-5.4 Mini 60%
  26. 26 Qwen3.6 27B 59.3%
  27. 27 Claude Sonnet 4.6 59.1%
  28. 28 Muse Spark 59%
  29. 29 GPT-5.3-Codex-Spark 58.4%
  30. 30 GPT-5.1-Codex 57.8%
  31. 31 Grok 4.20 reasoning 57.3%
  32. 32 MiMo-V2-Pro 57.1%
  33. 33 MiniMax-M2.7 57%
  34. 34 DeepSeek-V4-Flash 56.9%
  35. 35 GLM-5 56.2%
  36. 36 Seed2.0 Pro 55.8%
  37. 37 MAI-Code-1-Flash 54.8%
  38. 38 Tencent Hy3 preview 54.4%
  39. 39 Qwen3.5 397B-A17B 52.5%
  40. 40 MiniMax-M2.5 51.7%
  41. 41 Qwen 3.6 35B-A3B 51.5%
  42. 42 Step 3.5 Flash 51%
  43. 43 Kimi K2.5 50.8%
  44. 44 GPT-5 49.6%
  45. 45 Qwen3.5-122B-A10B 49.4%
  46. 46 MiniMax-M2.1 47.9%
  47. 47 Composer 1.5 47.9%
  48. 48 GPT-5.1 47.6%
  49. 49 Claude Sonnet 4.5 46.5%
  50. 50 DeepSeek-V3.2 46.4%
  51. 51 GPT-5.4 Nano 46.3%
  52. 52 MAI-Thinking-1 46%
  53. 53 Laguna M.1 45.8%
  54. 54 Seed1.8 45.2%
  55. 55 GPT-5-Codex 44.3%
  56. 56 Seed2.0 Lite (0428) 43.3%
  57. 57 Qwen3.5-27B 41.6%
  58. 58 GLM-4.7 41%
  59. 59 Qwen3.5-35B-A3B 40.5%
  60. 60 Ling-2.6-1T 40.5%
  61. 61 MiMo-V2-Flash 38.5%
  62. 62 Claude Opus 4.1 38%
  63. 63 Laguna XS 2.1 37.5%
  64. 64 Qwen3-Coder-Next 36.2%
  65. 65 North Mini Code 1.0 36%
  66. 66 Kimi K2 Thinking 35.7%
  67. 67 Laguna XS.2 35.7%
  68. 68 Claude Haiku 4.5 35.5%
  69. 69 GPT-5 mini 34.8%
  70. 70 Gemini 2.5 Pro Jun 2025 32.6%
  71. 71 Devstral 2 32.6%
  72. 72 Nemotron 3 Super 31%
  73. 73 MiniMax-M2 30%
  74. 74 K-EXAONE 29%
  75. 75 Kimi K2 Sep 2025 27.8%
  76. 76 Grok 4 27.2%
  77. 77 Qwen3-Coder-480B-A35B 27.2%
  78. 78 grok-code-fast-1 25.8%
  79. 79 GLM-4.6 24.5%
  80. 80 Devstral Small 2 22.5%
  81. 81 GPT-5 nano 21.8%
  82. 82 Nemotron-Cascade 2 30B-A3B 21.1%
  83. 83 gpt-oss-120b 18.7%
  84. 84 Gemini 2.5 Flash Jun 2025 17.1%
  85. 85 Solar Pro 3 (March 2026 update) 10.1%
  86. 86 Qwen3.5-9B 9.21%
  87. 87 gpt-oss-20b 3.37%

Sources: Official Terminal-Bench 2.0 leaderboard and labs’ model cards

#EntryScoreSettingDateSource
1 GPT-5.5 84.7% NexAU-AHE Apr 2026 Official leaderboard
2 Gemini 3.1 Pro 80.2% TongAgents Feb 2026 Official leaderboard
3 Claude Opus 4.7 80.2% WOZCODE Apr 2026 Official leaderboard
4 GPT-5.3 Codex 78.4% SageAgent Feb 2026 Official leaderboard
5 Claude Opus 4.6 76.4% Meta-Harness Feb 2026 Official leaderboard
6 Gemini 3 Pro 69.4% Ante Nov 2025 Official leaderboard
7 GPT-5.2 Codex 66.5% Deep Agents Dec 2025 Official leaderboard
8 GPT-5.2 64.9% Droid Dec 2025 Official leaderboard
9 Gemini 3 Flash 64.3% Junie CLI Dec 2025 Official leaderboard
10 Claude Opus 4.5 63.1% Droid Nov 2025 Official leaderboard
11 GPT-5.1-Codex-mini 61.6% hookele Nov 2025 Official leaderboard
12 GPT-5.1-Codex-Max 60.5% Codex CLI Nov 2025 Official leaderboard
13 GPT-5.1-Codex 57.8% Crux Nov 2025 Official leaderboard
14 Grok 4.20 57.3% reasoning, Grok CLI Feb 2026 Official leaderboard
15 Claude Sonnet 4.6 53.4% Simplai Agent Feb 2026 Official leaderboard
16 GLM-5 52.4% Terminus 2 Feb 2026 Official leaderboard
17 GPT-5 49.6% Codex CLI Aug 2025 Official leaderboard
18 GPT-5.1 47.6% Terminus 2 Nov 2025 Official leaderboard
19 Claude Sonnet 4.5 46.5% CAMEL-AI Sep 2025 Official leaderboard
20 MiniMax-M2.7 45.1% IndusAGI Coding Agent Mar 2026 Official leaderboard
21 GPT-5-Codex 44.3% Codex CLI Sep 2025 Official leaderboard
22 Kimi K2.5 43.2% Terminus 2 Jan 2026 Official leaderboard
23 MiniMax M2.7 Highspeed 42.9% Harness Agent Mar 2026 Official leaderboard
24 MiniMax-M2.5 42.7% cchuter Feb 2026 Official leaderboard
25 DeepSeek-V3.2 39.5% Terminus 2 Dec 2025 Official leaderboard
26 Claude Opus 4.1 38% Terminus 2 Aug 2025 Official leaderboard
27 MiniMax-M2.1 36.6% Crux Dec 2025 Official leaderboard
28 Kimi K2 Thinking 35.7% Terminus 2 Nov 2025 Official leaderboard
29 Claude Haiku 4.5 35.5% Goose Oct 2025 Official leaderboard
30 GPT-5 mini 34.8% spoox-o-m Aug 2025 Official leaderboard
31 GLM-4.7 33.4% Terminus 2 Dec 2025 Official leaderboard
32 Gemini 2.5 Pro Jun 2025 32.6% Terminus 2 Jun 2025 Official leaderboard
33 MiniMax-M2 30% Terminus 2 Oct 2025 Official leaderboard
34 Kimi K2 Sep 2025 27.8% Terminus 2 Sep 2025 Official leaderboard
35 Grok 4 27.2% OpenHands Jul 2025 Official leaderboard
36 Qwen3-Coder-480B-A35B 27.2% Dakou Agent Jul 2025 Official leaderboard
37 grok-code-fast-1 25.8% Mini-SWE-Agent Aug 2025 Official leaderboard
38 Qwen 3.6 35B-A3B 24.6% little-coder Apr 2026 Official leaderboard
39 GLM-4.6 24.5% Terminus 2 Sep 2025 Official leaderboard
40 GPT-5 nano 21.8% spoox-o-m Aug 2025 Official leaderboard
41 TermiGen-32B 19.3% Bash Agent May 2026 Official leaderboard
42 gpt-oss-120b 18.7% Terminus 2 Aug 2025 Official leaderboard
43 Gemini 2.5 Flash Jun 2025 17.1% Mini-SWE-Agent Jun 2025 Official leaderboard
44 AfterQuery-GPT-OSS-20B 17% Terminus 2 Mar 2026 Official leaderboard
45 Qwen3.5-9B 9.21% little-coder Feb 2026 Official leaderboard
46 gpt-oss-20b 3.37% Mini-SWE-Agent Aug 2025 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.5 84.7% NexAU-AHE Apr 2026 Official leaderboard
2 Claude Mythos Preview 82% Apr 2026 Lab model card
3 Gemini 3.1 Pro 80.2% TongAgents Feb 2026 Official leaderboard
4 Claude Opus 4.7 80.2% WOZCODE Apr 2026 Official leaderboard
5 GPT-5.3 Codex 78.4% SageAgent Feb 2026 Official leaderboard
6 Claude Opus 4.6 76.4% Meta-Harness Feb 2026 Official leaderboard
7 Qwen3.7-Plus 70.3% Jun 2026 Lab model card
8 Qwen3.7-Max 69.7% May 2026 Lab model card
9 Gemini 3 Pro 69.4% Ante Nov 2025 Official leaderboard
10 Composer 2.5 69.3% May 2026 Lab model card
11 GLM-5.1 69% Apr 2026 Lab model card
12 MiMo-V2.5-Pro 68.4% Apr 2026 Lab model card
13 DeepSeek-V4-Pro 67.9% Apr 2026 Lab model card
14 Kimi K2.6 66.7% Apr 2026 Lab model card
15 GPT-5.2 Codex 66.5% Deep Agents Dec 2025 Official leaderboard
16 MiMo-V2.5 65.8% Apr 2026 Lab model card
17 Qwen 3.6 Max (Preview) 65.4% Apr 2026 Lab model card
18 GPT-5.2 64.9% Droid Dec 2025 Official leaderboard
19 Gemini 3 Flash 64.3% Junie CLI Dec 2025 Official leaderboard
20 Claude Opus 4.5 63.1% Droid Nov 2025 Official leaderboard
21 Composer 2 61.7% Mar 2026 Lab model card
22 Qwen 3.6 Plus 61.6% Apr 2026 Lab model card
23 GPT-5.1-Codex-mini 61.6% hookele Nov 2025 Official leaderboard
24 GPT-5.1-Codex-Max 60.5% Codex CLI Nov 2025 Official leaderboard
25 GPT-5.4 Mini 60% Mar 2026 Lab model card
26 Qwen3.6 27B 59.3% Apr 2026 Lab model card
27 Claude Sonnet 4.6 59.1% Feb 2026 Lab model card
28 Muse Spark 59% Apr 2026 Lab model card
29 GPT-5.3-Codex-Spark 58.4% Feb 2026 Lab model card
30 GPT-5.1-Codex 57.8% Crux Nov 2025 Official leaderboard
31 Grok 4.20 57.3% reasoning, Grok CLI Feb 2026 Official leaderboard
32 MiMo-V2-Pro 57.1% Mar 2026 Lab model card
33 MiniMax-M2.7 57% Mar 2026 Lab model card
34 DeepSeek-V4-Flash 56.9% Apr 2026 Lab model card
35 GLM-5 56.2% Feb 2026 Lab model card
36 Seed2.0 Pro 55.8% Feb 2026 Lab model card
37 MAI-Code-1-Flash 54.8% Jun 2026 Lab model card
38 Tencent Hy3 preview 54.4% Apr 2026 Lab model card
39 Qwen3.5 397B-A17B 52.5% Feb 2026 Lab model card
40 MiniMax-M2.5 51.7% Feb 2026 Lab model card
41 Qwen 3.6 35B-A3B 51.5% Apr 2026 Lab model card
42 Step 3.5 Flash 51% Feb 2026 Lab model card
43 Kimi K2.5 50.8% Jan 2026 Lab model card
44 GPT-5 49.6% Codex CLI Aug 2025 Official leaderboard
45 Qwen3.5-122B-A10B 49.4% Feb 2026 Lab model card
46 MiniMax-M2.1 47.9% Dec 2025 Lab model card
47 Composer 1.5 47.9% Feb 2026 Lab model card
48 GPT-5.1 47.6% Terminus 2 Nov 2025 Official leaderboard
49 Claude Sonnet 4.5 46.5% CAMEL-AI Sep 2025 Official leaderboard
50 DeepSeek-V3.2 46.4% Dec 2025 Lab model card
51 GPT-5.4 Nano 46.3% Mar 2026 Lab model card
52 MAI-Thinking-1 46% Jun 2026 Lab model card
53 Laguna M.1 45.8% May 2026 Lab model card
54 Seed1.8 45.2% Dec 2025 Lab model card
55 GPT-5-Codex 44.3% Codex CLI Sep 2025 Official leaderboard
56 Seed2.0 Lite (0428) 43.3% Apr 2026 Lab model card
57 Qwen3.5-27B 41.6% Feb 2026 Lab model card
58 GLM-4.7 41% Dec 2025 Lab model card
59 Qwen3.5-35B-A3B 40.5% Feb 2026 Lab model card
60 Ling-2.6-1T 40.5% Jun 2026 Lab model card
61 MiMo-V2-Flash 38.5% Dec 2025 Lab model card
62 Claude Opus 4.1 38% Terminus 2 Aug 2025 Official leaderboard
63 Laguna XS 2.1 37.5% Jul 2026 Lab model card
64 Qwen3-Coder-Next 36.2% Feb 2026 Lab model card
65 North Mini Code 1.0 36% Jun 2026 Lab model card
66 Kimi K2 Thinking 35.7% Terminus 2 Nov 2025 Official leaderboard
67 Laguna XS.2 35.7% May 2026 Lab model card
68 Claude Haiku 4.5 35.5% Goose Oct 2025 Official leaderboard
69 GPT-5 mini 34.8% spoox-o-m Aug 2025 Official leaderboard
70 Gemini 2.5 Pro Jun 2025 32.6% Terminus 2 Jun 2025 Official leaderboard
71 Devstral 2 32.6% Dec 2025 Lab model card
72 Nemotron 3 Super 31% Mar 2026 Lab model card
73 MiniMax-M2 30% Terminus 2 Oct 2025 Official leaderboard
74 K-EXAONE 29% Dec 2025 Lab model card
75 Kimi K2 Sep 2025 27.8% Terminus 2 Sep 2025 Official leaderboard
76 Grok 4 27.2% OpenHands Jul 2025 Official leaderboard
77 Qwen3-Coder-480B-A35B 27.2% Dakou Agent Jul 2025 Official leaderboard
78 grok-code-fast-1 25.8% Mini-SWE-Agent Aug 2025 Official leaderboard
79 GLM-4.6 24.5% Terminus 2 Sep 2025 Official leaderboard
80 Devstral Small 2 22.5% Dec 2025 Lab model card
81 GPT-5 nano 21.8% spoox-o-m Aug 2025 Official leaderboard
82 Nemotron-Cascade 2 30B-A3B 21.1% Mar 2026 Lab model card
83 gpt-oss-120b 18.7% Terminus 2 Aug 2025 Official leaderboard
84 Gemini 2.5 Flash Jun 2025 17.1% Mini-SWE-Agent Jun 2025 Official leaderboard
85 Solar Pro 3 (March 2026 update) 10.1% Mar 2026 Lab model card
86 Qwen3.5-9B 9.21% little-coder Feb 2026 Official leaderboard
87 gpt-oss-20b 3.37% Mini-SWE-Agent Aug 2025 Official leaderboard

Labs report scores for 63 models in their cards. Epoch AI collected claims for 5 models from labs’ reports and papers. Switch the scores above to rank them together.

6 models have held number one on the official leaderboard since its release in November 2025 7 models have held number one with claims included since its release in November 2025

Lab claims count from the day the card came out, independent runs from the model’s release. We draw one row each for the benchmark’s own leaderboard and what labs claim in their cards.

OpenAI Google DeepMind Anthropic
Own leaderboard Replaced by 2.1, May 2026
GPT-5.5 84.7%
Lab claims Replaced by 2.1, Jul 2026
GPT-5.5 82.7%

Sources: Official Terminal-Bench 2.0 leaderboard and labs’ model cards

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Reported by the lab in its model card A claim Epoch AI collected from a report or paper Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5.5 Gemini 3.1 Pro Gemini 3 Pro GPT-5.1-Codex-mini GPT-5 Claude Opus 4.1 Gemini 2.5 Pro 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5.5 GPT-5.3 Codex

Sources: Official Terminal-Bench 2.0 leaderboard, labs’ model cards and claims collected by Epoch AI (CC BY 4.0)

4 labs put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Google DeepMind
OpenAI
Anthropic
DeepSeek
Mistral AI
Xiaomi Corp
ByteDance Seed
Z.ai
MiniMax
LG AI Research
Moonshot AI
StepFun
Qwen
Cursor
NVIDIA
Upstage
Meta
Tencent
Poolside
Microsoft
Cohere
DateLaunch postScoreAs the post writes it
Jul 2, 2026 Laguna XS 2.1 37.5 Terminal-Bench 2.0
Jun 9, 2026 North Mini Code 1.0 36 Terminal-Bench v2 (simple ReAct harness with one terminal tool)
Jun 2, 2026 MAI-Code-1-Flash 54.8 Terminal Bench 2
Jun 2, 2026 MAI-Thinking-1 46 Terminal-Bench 2.0
Jun 1, 2026 Qwen3.7-Plus 70.3 Terminal Bench 2.0-Terminus (avg of 5 runs)
May 20, 2026 Qwen3.7-Max 69.7 Terminal Bench 2.0-Terminus
May 18, 2026 Composer 2.5 69.3 Terminal-Bench 2.0
Apr 28, 2026 Laguna XS.2 30.1 Terminal-Bench 2.0 (mean pass@1 over 5 runs)
Apr 28, 2026 Laguna M.1 40.7 Terminal-Bench 2.0 (mean pass@1 over 5 runs, 48GB RAM/32 CPUs)
Apr 28, 2026 Seed2.0 Lite (0428) 43.3 Terminal Bench 2.0
Apr 27, 2026 MiMo-V2.5-Pro 68.4 Terminal-Bench 2.0
Apr 24, 2026 DeepSeek-V4-Flash 56.9 Terminal Bench 2.0 (Acc), V4-Flash Max
Apr 24, 2026 DeepSeek-V4-Pro 67.9 Terminal Bench 2.0 (Acc)
Apr 23, 2026 Tencent Hy3 preview 54.4 Terminal-Bench 2.0
Apr 23, 2026 GPT-5.5 82.7 Terminal-Bench 2.0
Apr 22, 2026 Qwen3.6 27B 59.3 Terminal-Bench 2.0 (Terminus-2, avg of 5)
Apr 22, 2026 MiMo-V2.5 65.8 Terminal-Bench 2.0
Apr 20, 2026 Qwen 3.6 Max (Preview) 65.4 Terminal-Bench 2.0 (Terminus-2) Agentic Terminal Coding
Apr 20, 2026 Kimi K2.6 66.7 Terminal-Bench 2.0 (Terminus-2)
Apr 16, 2026 Claude Opus 4.7 69.4 Agentic terminal coding Terminal-Bench 2.0
Apr 15, 2026 Qwen 3.6 35B-A3B 51.5 Terminal-Bench 2.0 (Terminus-2, avg of 5)
Apr 8, 2026 Muse Spark 59 Terminal-Bench 2.0 (Agentic Terminal Coding)
Apr 7, 2026 GLM-5.1 69 Terminal-Bench 2.0 (Best self-reported), Claude Code
Apr 7, 2026 Claude Mythos Preview 82 Terminal-Bench 2.0 (Terminus-2)
Apr 2, 2026 Qwen 3.6 Plus 61.6 Terminal-Bench 2.0
Mar 24, 2026 Solar Pro 3 (March 2026 update) 10.1 Terminal Bench 2
Mar 19, 2026 Nemotron-Cascade 2 30B-A3B 21.1 Terminal Bench 2.0
Mar 19, 2026 Composer 2 61.7 Terminal-Bench 2.0 (Harbor, mean of 5 runs)
Mar 18, 2026 MiMo-V2-Pro 57.1 Terminal-Bench 2.0
Mar 18, 2026 MiniMax-M2.7 57 Terminal Bench 2 (text)
Mar 17, 2026 GPT-5.4 Nano 46.3 Terminal-Bench 2.0
Mar 17, 2026 GPT-5.4 Mini 60 Terminal-Bench 2.0
Mar 11, 2026 Nemotron 3 Super 31 Terminal Bench Core 2.0
Feb 24, 2026 Qwen3.5-35B-A3B 40.5 Terminal Bench 2
Feb 24, 2026 Qwen3.5-27B 41.6 Terminal Bench 2
Feb 24, 2026 Qwen3.5-122B-A10B 49.4 Terminal Bench 2
Feb 19, 2026 Gemini 3.1 Pro 68.5 Terminal-Bench 2.0 (Terminus-2 harness)
Feb 17, 2026 Claude Sonnet 4.6 59.1 Agentic terminal coding Terminal-Bench 2.0
Feb 16, 2026 Qwen3.5 397B-A17B 52.5 Terminal Bench 2
Feb 14, 2026 Seed2.0 Pro 55.8 Terminal Bench 2.0 (Terminus 2, 3 tasks excluded)
Feb 12, 2026 GPT-5.3-Codex-Spark 58.4 Terminal-Bench 2.0
Feb 12, 2026 MiniMax-M2.5 51.7 Terminal Bench 2
Feb 11, 2026 GLM-5 56.2 Terminal-Bench 2.0 (Terminus 2)
Feb 9, 2026 Composer 1.5 47.9 Terminal-Bench 2.0 Score (%) (Harbor, mean of 2 runs)
Feb 5, 2026 Claude Opus 4.6 65.4 Agentic terminal coding Terminal-Bench 2.0
Feb 5, 2026 GPT-5.3 Codex 77.3 Terminal-Bench 2.0
Feb 2, 2026 Qwen3-Coder-Next 36.2 Terminal-Bench 2.0 (w/ Terminus-2 json)
Feb 2, 2026 Step 3.5 Flash 51 Terminal-Bench 2.0
Jan 27, 2026 Kimi K2.5 50.8 Terminal Bench 2.0
Dec 31, 2025 K-EXAONE 29 Terminal-Bench 2.0
Dec 23, 2025 MiniMax-M2.1 47.9 Terminal-bench 2.0
Dec 22, 2025 GLM-4.7 41 Terminal Bench 2.0
Dec 18, 2025 GPT-5.2 Codex 64 Terminal-Bench 2.0
Dec 18, 2025 Seed1.8 45.2 Terminal Bench 2.0
Dec 17, 2025 Gemini 3 Flash 47.6 Terminal-bench 2.0 (Terminus-2 harness)
Dec 16, 2025 MiMo-V2-Flash 38.5 Terminal-Bench 2.0
Dec 9, 2025 Devstral Small 2 22.5 Terminal Bench 2 (HF card)
Dec 9, 2025 Devstral 2 32.6 Terminal Bench 2 (HF card)
Dec 1, 2025 DeepSeek-V3.2 46.4 Terminal Bench 2.0 (Acc)
Nov 24, 2025 Claude Opus 4.5 59.3 Agentic terminal coding Terminal-bench 2.0 (128k thinking budget)
Nov 19, 2025 GPT-5.1-Codex-Max 58.1 Terminal-Bench 2.0, xhigh
Nov 18, 2025 Gemini 3 Pro 54.2 Terminal-Bench 2.0 (Terminus-2 agent)

28 system cards report it, from 17 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Jun 13, 2026 Ling-2.6-1T technical report 40.5 terminal-bench 2.0, accuracy (Table 5) Capabilities
Jun 12, 2026 Muse Spark system card 59 Terminal-Bench 2.0 (2.3.1.1 AI R&D text) AI R&D risk
Jun 2, 2026 MAI-Thinking-1 technical report 46 Terminal-Bench 2.0, bash tool only, timeouts ignored (Table 11) Capabilities
Jun 2, 2026 MAI-Code-1-Flash model card 54.8 Terminal Bench 2, pass rate (Coding table) Capabilities
May 26, 2026 MiniMax-M2.7 technical report 57 Terminal-Bench 2.0, Terminus-2, verified 2.0 dataset (zai-org/terminal-bench-2-verified), 4 trials (Table 4) Capabilities
May 26, 2026 MiniMax-M2.5 technical report 51.7 Terminal-Bench 2.0, Terminus-2, verified 2.0 dataset (zai-org/terminal-bench-2-verified), 4 trials (Table 4) Capabilities
May 25, 2026 Laguna XS.2 technical report 35.7 Terminal-Bench 2.0, pool harness, 500 steps, mean pass@1 over 4 runs, some task images and verifiers patched (Table 6) Capabilities
May 25, 2026 Laguna M.1 technical report 45.8 Terminal-Bench 2.0, pool harness, 500 steps, mean pass@1 over 4 runs, some task images and verifiers patched (Table 7) Capabilities
Apr 26, 2026 DeepSeek-V4-Pro technical report 67.9 Terminal Bench 2.0 (Acc), original dataset, DS-V4-Pro Max (Table 6). About 72.0 on the Verified subset (5.3.1 text) Capabilities
Apr 26, 2026 DeepSeek-V4-Flash technical report 56.9 Terminal Bench 2.0 (Acc), DeepSeek-V4-Flash Max (Table 7) Capabilities
Apr 16, 2026 Claude Opus 4.7 system card 69.4 Terminal-Bench 2.0, Terminus-2 harness, thinking disabled, mean reward over 5 attempts on 89 tasks (Table 8.1.A, 8.3) Capabilities
Apr 7, 2026 Claude Mythos Preview system card 82 Terminal-Bench 2.0, Terminus-2 harness in Harbor, adaptive thinking, max effort, 1x timeouts, mean reward over 5 attempts on 89 tasks (Table 6.3.A, 6.5); 92.1 with the 2.1 fixes and 4-hour timeouts Capabilities
Mar 25, 2026 Composer 2 technical report 61.7 Terminal-Bench, Cursor's own harness, version not stated, cites the ICLR 2026 Terminal-Bench paper (Table 1 and 7.2) Capabilities
Mar 21, 2026 Seed1.8 model card 45.2 Terminal Bench 2.0 (Table 4) Capabilities
Mar 19, 2026 Nemotron-Cascade 2 30B-A3B technical report 21.1 Terminal Bench 2.0, Terminus-2, avg@5 (Table 1) Capabilities
Mar 11, 2026 Nemotron 3 Super technical report 31 Terminal Bench Core 2.0, Harbor harness (Table 5) Capabilities
Feb 28, 2026 Qwen3-Coder-Next technical report 36.2 Terminal-Bench 2.0, Terminus2-json, as in Figure 1. Terminus2-xml 34.2, ClaudeCode 30.9, QwenCode 25.8 (Table 5) Capabilities
Feb 19, 2026 Gemini 3.1 Pro model card 68.5 Terminal-Bench 2.0, Terminus-2 harness (results table, p.4) Capabilities
Feb 17, 2026 GLM-5 technical report 56.2 Terminal-Bench 2.0 (Terminus-2), original dataset (Table 7). Verified dataset 60.7; Claude Code 56.2 / 61.1 Capabilities
Feb 17, 2026 Claude Sonnet 4.6 system card 59.1 Terminal-Bench 2.0, max effort, no thinking budget, Terminus-2 harness (Table 2.1.A, 2.3) Capabilities
Feb 14, 2026 Seed2.0 Pro model card 55.8 Terminal Bench 2.0, Terminus 2, 3 tasks excluded (Table 11) Capabilities
Feb 11, 2026 Step 3.5 Flash technical report 51 Terminal-Bench 2.0, avg@8 50.98%, pass@8 67/89 (Table 5 and E.2.2) Capabilities
Feb 5, 2026 Claude Opus 4.6 system card 65.4 Terminal-Bench 2.0, Terminus-2 harness, adaptive thinking, max effort, 89 tasks x 15 runs (Table 2.3.A, Figure 2.5.A); 61.1 at medium and 55.1 at low effort Capabilities
Jan 30, 2026 Kimi K2.5 technical report 50.8 Terminal Bench 2.0, Terminus-2, non-thinking mode (Table 4) Capabilities
Jan 6, 2026 MiMo-V2-Flash technical report 38.5 Terminal Bench 2.0 (Table 9) Capabilities
Jan 5, 2026 K-EXAONE technical report 29 Terminal-Bench 2.0, Terminus 2 agent (Table 3, Reasoning mode) Capabilities
Dec 1, 2025 DeepSeek-V3.2 technical report 46.4 Terminal Bench 2.0 (Acc), Claude Code framework, DeepSeek-V3.2 Thinking (Table 2). 39.3 with Terminus in non-thinking mode Capabilities
Nov 24, 2025 Claude Opus 4.5 system card 59.3 Terminal-bench 2.0, Terminus-2 harness, 128k thinking budget (Table 2.3.A; 57.8% with 64k thinking, 2.5) Capabilities

The lab's number comes out ahead in 9 of 19 pairs that differ

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. Every benchmark.

Model Lab says Board says Gap What differs
Qwen 3.6 35B-A3B 51.5% Terminal-Bench 2.0 (Terminus-2, avg of 5) 24.6% Leaderboard, little-coder +27 pts Qwen uses Terminus-2 with a three-hour timeout, while the compared leaderboard entry uses little-coder.
Gemini 3 Flash 47.6% Terminal-bench 2.0 (Terminus-2 harness) 64.3% Leaderboard, Junie CLI −17 pts Google uses the Terminus-2 harness, while the board pairs Gemini 3 Flash with Junie CLI.
Gemini 3 Pro 54.2% Terminal-Bench 2.0 (Terminus-2 agent) 69.4% Leaderboard, Ante −15 pts The lab used Terminus-2, while the board reports the Ante agent.
MiniMax-M2.7 57% Terminal Bench 2 (text) 45.1% Leaderboard, IndusAGI Coding Agent +12 pts
Gemini 3.1 Pro 68.5% Terminal-Bench 2.0 (Terminus-2 harness) 80.2% Leaderboard, TongAgents −12 pts
MiniMax-M2.1 47.9% Terminal-bench 2.0 36.6% Leaderboard, Crux +11 pts
Claude Opus 4.6 65.4% Agentic terminal coding Terminal-Bench 2.0 76.4% Leaderboard, Meta-Harness −11 pts
Claude Opus 4.7 69.4% Agentic terminal coding Terminal-Bench 2.0 80.2% Leaderboard, WOZCODE −11 pts
MiniMax-M2.5 51.7% Terminal Bench 2 42.7% Leaderboard, cchuter +9.0 pts
GLM-4.7 41% Terminal Bench 2.0 33.4% Leaderboard, Terminus 2 +7.6 pts
Kimi K2.5 50.8% Terminal Bench 2.0 43.2% Leaderboard, Terminus 2 +7.6 pts
DeepSeek-V3.2 46.4% Terminal Bench 2.0 (Acc) 39.5% Leaderboard, Terminus 2 +6.9 pts
Claude Sonnet 4.6 59.1% Agentic terminal coding Terminal-Bench 2.0 53.4% Leaderboard, Simplai Agent +5.7 pts
Claude Opus 4.5 59.3% Agentic terminal coding Terminal-bench 2.0 (128k thinking budget) 63.1% Leaderboard, Droid −3.9 pts
GLM-5 56.2% Terminal-Bench 2.0 (Terminus 2) 52.4% Leaderboard, Terminus 2 +3.8 pts
GPT-5.2 Codex 64% Terminal-Bench 2.0 66.5% Leaderboard, Deep Agents −2.5 pts
GPT-5.1-Codex-Max 58.1% Terminal-Bench 2.0, xhigh 60.5% Leaderboard, Codex CLI −2.4 pts
GPT-5.5 82.7% Terminal-Bench 2.0 84.7% Leaderboard, NexAU-AHE −2.0 pts
GPT-5.3 Codex 77.3% Terminal-Bench 2.0 78.4% Leaderboard, SageAgent −1.1 pts

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

Google DeepMind Anthropic OpenAI
Terminal-Bench 4.1 20 leaders
Claude Opus 5.5 65.8%
Terminal-Bench Replaced by 2.0, Nov 2025
GPT-5 49.6%
Terminal-Bench 2.0 Replaced by 2.1, May 2026
GPT-5.5 84.7%
Terminal-Bench 2.1 Passed 85% in Sep 2026, by GPT-6 Astra
GPT-6 Astra 87.4%
Terminal-Bench 3.0 Replaced by 4.0, Aug 2026
Claude Opus 5 42.7%
Terminal-Bench 4.0 4 leaders
Claude Opus 5.5 64.8%
Terminal-Bench 4.1 1 leader
Claude Opus 5.5 65.8%

Sources: Official leaderboards and boards copied by Epoch AI (CC BY 4.0)

Replaces Terminal-Bench Replaced by Terminal-Bench 2.1

More in Terminal agents

All 19 in Terminal agents
  • SkillsBench In use 86-87 containerized expert tasks in 11 domains, run with and without Agent Skills. Verifier tests give pass rate and skill lift. GPT-5.5 67.3%
  • Terminal-Bench Hard In use Agents complete 44 hard tasks from Terminal-Bench Core in the Terminus 2 harness. Artificial Analysis scores pass@1 averaged over 3 repeats. Gemma 4 31B IT 36%
  • Harbor-Index In use Agents solve about 80 hard tasks drawn from 29 benchmarks run through Harbor adapters, such as HLE, GSO and SWE-bench Pro. Each task keeps its source verifier. GPT-5.5 28.1%
  • Terminal-Bench 2.1 Extended Not in model cards Agents complete long command-line tasks in containers, scored by verification tests. Mercor adds 99 new expert-written tasks to the 89 public ones and reports both scores. Muse Wonder 54%
  • o11y-bench Not in model cards Agents query metrics, logs and traces in a live Grafana stack, investigate incidents and fix dashboards. Grading checks claims against facts read from the same environment. Claude Opus 5.5 93.8%
  • CompileBench Not in model cards Agents build real open-source projects through dependency problems, legacy toolchains and complex build systems, for example compiling 22-year-old code. Claude Opus 4.1 100%