GameBench 2 by reasoning setting leaderboard
As of 4 Nov 2026, the top score on GameBench 2 by reasoning setting is 94.8 by Claude Opus 5.5 (Anthropic), ahead of Claude Fable 5.1 at 94.1, across 186 model configurations with a published result (DuelLab GameBench 2, result dated 3 Nov 2026). GameBench 2 scored separately for each reasoning setting a model was run at, filed under the matching effort level. Only settings DuelLab ranks are shown. GameBench 2 by reasoning setting is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 4 Nov 2026)
- Claude Opus 5.5 holds the top GameBench 2 by reasoning setting score at 94.8 (DuelLab GameBench 2, 3 Nov 2026).
- Claude Fable 5.1 is second on GameBench 2 by reasoning setting at 94.1 (DuelLab GameBench 2, 3 Nov 2026).
- GameBench 2 by reasoning setting: GameBench 2 scored separately for each reasoning setting a model was run at, filed under the matching effort level. Only settings DuelLab ranks are shown (DuelLab GameBench 2).
- 186 model configurations have a published GameBench 2 by reasoning setting result on BenchLeader (BenchLeader, 4 Nov 2026).
- GameBench 2 by reasoning setting does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
GameBench 2 scored separately for each reasoning setting a model was run at, filed under the matching effort level. Only settings DuelLab ranks are shown.
- Published by
- DuelLab GameBench 2
- Category
- Coding
- Index weight
- Reference only
- Models
- 186
- Data as of
- 4 Nov 2026
DuelLab GameBench 2, release gb2-20261103-09b2218cbd85 (https://benchmarks.duellab.org/releases/gb2-20261103-09b2218cbd85/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
- 1Claude Opus 5.5 (xhigh)94.8
- 2Claude Fable 5.1 (xhigh)94.1
- 3GPT-6 Astra (xhigh)94.0
- 4GPT-6.1 Sol (xhigh)92.0
- 5Claude Opus 5.5 (medium)91.5
- 6Claude Sonnet 5.5 (xhigh)91.3
- 7Gemini 4 Argon (high)89.0
- 8GPT-6 Astra (medium)88.6
- 9GPT-6.1 Sol (medium)86.3
- 10Claude Fable 5.1 (low)84.8
- 11Claude Opus 5 (xhigh)84.7
- 12Kimi K3.1 Pro (max)84.0
- 13GPT-6 Astra (low)83.5
- 14Claude Fable 5.1 (medium)82.9
- 15GPT-5.6 Sol (xhigh)79.6
186 of 186
| # | ||||
|---|---|---|---|---|
| 1 |  Claude Opus 5.5xhigh | 94.8 | 66.2 | 2026-11-03 |
| 2 | Claude Fable 5.1xhigh | 94.1 | 65.2 | 2026-11-03 |
| 3 |  GPT-6 Astraxhigh | 94.0 | 66.7 | 2026-11-03 |
| 4 | GPT-6.1 Solxhigh | 92.0 | 64.5 | 2026-11-03 |
| 5 | Claude Opus 5.5medium | 91.5 | 61.9 | 2026-11-03 |
| 6 | Claude Sonnet 5.5xhigh | 91.3 | 63.4 | 2026-11-03 |
| 7 |  Gemini 4 Argonhigh | 89.0 | 66.0 | 2026-11-03 |
| 8 | GPT-6 Astramedium | 88.6 | 65.0 | 2026-11-03 |
| 9 | GPT-6.1 Solmedium | 86.3 | 63.5 | 2026-11-03 |
| 10 | Claude Fable 5.1low | 84.8 | 59.1 | 2026-11-03 |
| 11 | Claude Opus 5xhigh | 84.7 | 60.3 | 2026-11-03 |
| 12 |  Kimi K3.1 Promax | 84.0 | – | 2026-11-03 |
| 13 | GPT-6 Astralow | 83.5 | 62.6 | 2026-11-03 |
| 14 | Claude Fable 5.1medium | 82.9 | 62.2 | 2026-11-03 |
| 15 | GPT-5.6 Solxhigh | 79.6 | 64.2 | 2026-11-03 |
| 16 | GPT-6 Solxhigh | 79.3 | 58.7 | 2026-11-03 |
| 17 | Gemini 4 Argonlow | 79.2 | – | 2026-11-03 |
| 18 | Kimi K3.1 Prolow | 79.2 | – | 2026-11-03 |
| 19 |  Muse Wonderxhigh | 77.1 | – | 2026-11-03 |
| 20 | Gemini 4 Argonmedium | 76.7 | – | 2026-11-03 |
| 21 | Claude Opus 5.5low | 76.3 | 59.4 | 2026-11-03 |
| 22 | GPT-6.1 Sollow | 76.1 | 60.7 | 2026-11-03 |
| 23 | Claude Opus 5medium | 75.1 | 57.4 | 2026-11-03 |
| 24 | Claude Fable 5xhigh | 75.0 | – | 2026-11-03 |
| 25 | GPT-6 Solmedium | 74.5 | 57.2 | 2026-11-03 |
| 26 |  GLM 5.3maxopen ↗ | 72.1 | 58.9 | 2026-11-03 |
| 27 |  Grok 4.7medium | 72.0 | – | 2026-11-03 |
| 28 | Grok 4.7xhigh | 71.5 | 58.2 | 2026-11-03 |
| 29 | Kimi K3.1 Prohigh | 71.5 | – | 2026-11-03 |
| 30 | Claude Opus 5no reasoning | 69.9 | – | 2026-11-03 |
| 31 | GLM 5.3highopen ↗ | 69.8 | – | 2026-11-03 |
| 32 |  DeepSeek V4.1 Flash 1017maxopen ↗ | 69.8 | 59.1 | 2026-11-03 |
| 33 | Muse Spark 1.3xhigh | 68.7 | 58.5 | 2026-11-03 |
| 34 | Grok 4.6medium | 68.6 | 59.6 | 2026-11-03 |
| 35 |  Qwen3.8 Maxxhigh | 68.1 | 57.1 | 2026-11-03 |
| 36 | GLM 5.3 FlashXmaxopen | 67.6 | – | 2026-11-03 |
| 37 | DeepSeek V4.1 Flashmaxopen ↗ | 67.4 | 59.0 | 2026-11-03 |
| 38 | Qwen3.8 2.4T A95Bxhighopen ↗ | 66.7 | – | 2026-11-03 |
| 39 | GPT-5.6 Solmedium | 65.8 | 61.5 | 2026-11-03 |
| 40 | Grok 4.6xhigh | 64.8 | 58.0 | 2026-11-03 |
| 41 | GPT-6 Solno reasoning | 64.8 | 49.2 | 2026-11-03 |
| 42 | Muse Spark 1.3medium | 64.7 | – | 2026-11-03 |
| 43 | DeepSeek V4.1 Flash 1017highopen ↗ | 64.3 | 51.4 | 2026-11-03 |
| 44 | Claude Fable 5medium | 64.3 | – | 2026-11-03 |
| 45 | Gemini 3.8 Flashhigh | 64.2 | 61.9 | 2026-11-03 |
| 46 | Ember-1max | 62.9 | – | 2026-11-03 |
| 47 | Gemini 3.6 Flashhigh | 62.8 | 61.3 | 2026-11-03 |
| 48 | Gemini 3.5 Flashhigh | 62.6 | 62.6 | 2026-11-03 |
| 49 | Claude Opus 4.8xhigh | 62.4 | – | 2026-11-03 |
| 50 | GPT-5.4xhigh | 61.8 | 63.3 | 2026-11-03 |
| 51 | Muse Wondermedium | 61.5 | – | 2026-11-03 |
| 52 | Grok 4.5high | 61.5 | 56.9 | 2026-11-03 |
| 53 | GLM 5.3 Flashhighopen ↗ | 61.4 | – | 2026-11-03 |
| 54 | Grok 4.5medium | 60.1 | – | 2026-11-03 |
| 55 | GPT-5.5xhigh | 59.7 | 64.4 | 2026-11-03 |
| 56 | Gemini 3.8 Flashmedium | 59.3 | 60.6 | 2026-11-03 |
| 57 | Muse Spark Openxhighopen ↗ | 59.2 | 53.8 | 2026-11-03 |
| 58 | Claude Fable 5low | 58.1 | – | 2026-11-03 |
| 59 | GLM 5.3 Flashmaxopen ↗ | 57.5 | 54.2 | 2026-11-03 |
| 60 | Muse Spark 1.2xhigh | 57.4 | 61.7 | 2026-11-03 |
More coding benchmarks · All LLM benchmarks
Related benchmarks
Top head-to-heads on GameBench 2 by reasoning setting
Cite as: BenchLeader, “GameBench 2 by reasoning setting leaderboard”, data as of 4 Nov 2026, https://www.benchleader.com/benchmarks/duellab_gb2_setting.
GameBench 2 by reasoning setting: questions
- What does GameBench 2 by reasoning setting measure?
- GameBench 2 scored separately for each reasoning setting a model was run at, filed under the matching effort level. Only settings DuelLab ranks are shown. Scores are reported in index; higher is better.
- What is the top score on GameBench 2 by reasoning setting?
- As of 4 Nov 2026, the highest score on GameBench 2 by reasoning setting is 94.8 by Claude Opus 5.5 (result dated 3 Nov 2026), ahead of Claude Fable 5.1 at 94.1. Scores come from DuelLab GameBench 2 and are re-read daily.
- Which AI model leads GameBench 2 by reasoning setting?
- Claude Opus 5.5 leads GameBench 2 by reasoning setting with 94.8 as of 4 Nov 2026 (result dated 3 Nov 2026), ahead of Claude Fable 5.1 at 94.1.
- How many models have GameBench 2 by reasoning setting results?
- 186 model configurations have a GameBench 2 by reasoning setting result on BenchLeader, all taken from DuelLab GameBench 2.
- Who runs GameBench 2 by reasoning setting and how often is it updated?
- GameBench 2 by reasoning setting is published by DuelLab GameBench 2. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does GameBench 2 by reasoning setting count toward the BenchLeader Index?
- No. GameBench 2 by reasoning setting is shown for reference but left out of the composite index.