activeCodingAgents

Terminal-Bench 2.0

Metric: % Tasks Completed (higher is better)Introduced: 2026

Second-generation terminal agent benchmark with 89 high-quality tasks spanning software engineering, machine learning, security, data science, and other real shell environments.

Models ranked

29

tracked on this benchmark

Score band

82.7 – 10.1

best → lowest tracked

Snapshot trend

-1.00

Jun 18 → Oct 17 · 1 models

Leaderboard

Tracked models ranked by % Tasks Completed (higher is better).

Compare candidates
#Model variant and provenanceScore
1
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
82.7
2
Version: Terminal-Bench 2.0; standard GPT-5.5 weightsHarness: Not recordedEvaluator: Not recordedObserved: Apr 23, 2026Confidence: Not recordedSource

Notes: Not directly comparable to Claude Opus 4.8; no Opus 4.8 Terminal-Bench 2.0 row was found. Use Terminal-Bench 2.1 for the head-to-head comparison.

82.7
3
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
77.3
4
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
76.2
5
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
75.1
6
Version: TerminusHarness: Not recordedEvaluator: Not recordedObserved: May 20, 2026Confidence: Not recordedSource
69.7
7
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
69.4
8
Version: Terminal-Bench 2.0Harness: Not recordedEvaluator: Not recordedObserved: May 21, 2026Confidence: Not recordedSource

Notes: Confidence: medium. secondary-source score for the Cursor IDE-native Composer 2.5 agent; useful for comparison but still reflects the full Cursor agent system rather than a standalone base-model run.

69.3
9
Version: Terminal-Bench 2.0 (score)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
68.4
10
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
66.7
11
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
66.0
12
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
65.4
13
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
65.4
14
Version: Terminal-Bench 2.0; OpenAI launch comparison; GPT-5.2-Codex xhighHarness: Not recordedEvaluator: Not recordedObserved: Dec 18, 2025Confidence: Not recordedSource

Notes: OpenAI GPT-5.3-Codex appendix reports GPT-5.2-Codex 64.0%, GPT-5.2 62.2%, GPT-5.1-Codex-Max 58.1%. Evaluator: OpenAI self-reported launch benchmark. Harness details are not fully reproducible from public text; compare separately from tbench.ai third-party agent rows. Confidence: confirmed for reported value, medium for reproducibility.

64.0
15
Version: Not recordedHarness: Not recordedEvaluator: Not recordedObserved: Apr 7, 2026Confidence: Not recordedSource
63.5
16
Version: Official Harbor evaluation framework (score)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
61.7
17
Version: Terminal-Bench 2.0Harness: Not recordedEvaluator: Not recordedObserved: Feb 17, 2026Confidence: Not recordedSource

Notes: Confidence: medium. DataCamp citing Anthropic release notes; terminal-bench.com leaderboard was not independently accessible, and Composer 2.5 comparisons remain cross-harness.

59.1
18
Version: independent evaluation (BenchLM, June 2026)Harness: Not recordedEvaluator: Not recordedObserved: Jun 18, 2026Confidence: Not recordedSource
59.1
19

Configuration: GPT-5.1-Codex-Max (xhigh, compaction)

Version: Terminal-Bench 2.0Harness: Codex CLI in Laude Institute Harbor harnessEvaluator: OpenAIObserved: Oct 17, 2026Confidence: self-reportedSource

Notes: OpenAI launch post (2025-11-19): xhigh reasoning effort with compaction; Codex CLI run inside the Laude Institute Harbor harness. (tbench.ai leaderboard shows 60.4 for Codex CLI + this model; not used, mirror-only.) Retrieved 2026-10-17.

58.1
20
Version: OpenRouter Terminal Bench 2 listingHarness: Not recordedEvaluator: Not recordedObserved: Jun 1, 2026Confidence: Not recordedSource
57.0
21
Version: Terminal Bench 2.0 accuracy, V4-Flash Think MaxHarness: Not recordedEvaluator: Not recordedObserved: Jul 3, 2026Confidence: Not recordedSource
56.9
22
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
54.8
23
Version: Terminal-Bench 2.0 (score)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
54.4
24
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
52.5
25
Version: Terminal-Bench 2.0 (accuracy%)Harness: Not recordedEvaluator: Not recordedObserved: Jun 7, 2026Confidence: Not recordedSource
50.8
Showing the top 25 of 29 tracked models. Browse all models.

How to read this benchmark

This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.

Trust this score when

  • There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
  • The model list covers the same version family you can actually deploy today.
  • Top candidates overlap with your required routing and feature requirements.

Be cautious when

  • There is only one benchmark snapshot or the dataset appears stale.
  • The benchmark metric direction is opposite of your decision objective.
  • The score difference between options is narrow and likely within implementation variance.

Related benchmarks

Last reviewed: May 21, 2026

Resources