GBAEval
ReportedNo description yet: we only describe an eval set when we can do it correctly.
- Kind
- Published by a third party's leaderboard, collected by Epoch AI. The leaderboard is named and linked.
- Score shown
- Epoch's column "Overall score", as a percentage, higher is better. The source gives no error bars.
- Results
- 28 in all: 7 for open-weight models, 21 for closed ones. 7 linked to a model page here.
Best open-weight result: moonshotai/Kimi-K3, 48.3%, number 14 of 28 results.
| # | Model | Weights | Setting | Score | Date | Source |
|---|---|---|---|---|---|---|
| 1 | claude-opus-5-5_unknownAnthropic | closed | effort unknown | 90.6% | 14 Nov 2026 | gbaeval.com |
| 2 | gpt-6-astra_unknownOpenAI | closed | effort unknown | 87.1% | 14 Nov 2026 | gbaeval.com |
| 3 | claude-fable-5-2_unknownAnthropic | closed | effort unknown | 84.8% | 14 Nov 2026 | gbaeval.com |
| 4 | gpt-6.1-sol_unknownOpenAI | closed | effort unknown | 81.9% | 14 Nov 2026 | gbaeval.com |
| 5 | claude-opus-5_unknownAnthropic | closed | effort unknown | 79.6% | 26 Jul 2026 | gbaeval.com |
| 6 | gemini-4-argon_unknownGoogle DeepMind | closed | effort unknown | 78.7% | 14 Nov 2026 | gbaeval.com |
| 7 | claude-fable-5_unknownAnthropic | closed | effort unknown | 74.5% | 10 Jun 2026 | gbaeval.com |
| 8 | claude-opus-4-8_unknownAnthropic | closed | effort unknown | 70.9% | 30 May 2026 | gbaeval.com |
| 9 | grok-4.5_unknownxAI | closed | effort unknown | 65.4% | 10 Jul 2026 | gbaeval.com |
| 10 | claude-sonnet-5_unknownAnthropic | closed | effort unknown | 65.3% | 13 Jul 2026 | gbaeval.com |
| 11 | gpt-5.5_unknownOpenAI | closed | effort unknown | 53.2% | 3 May 2026 | gbaeval.com |
| 12 | gpt-5.6-sol_unknownOpenAI | closed | effort unknown | 52.6% | 15 Jul 2026 | gbaeval.com |
| 13 | claude-sonnet-4-6_unknownAnthropic | closed | effort unknown | 48.8% | 3 May 2026 | gbaeval.com |
| 14 | moonshotai/Kimi-K3Epoch: kimi-k3_unknown, Moonshot | open | effort unknown | 48.3% | 20 Jul 2026 | gbaeval.com |
| 15 | gpt-5.4-2026-03-05_unknownOpenAI | closed | effort unknown | 45.1% | 12 May 2026 | gbaeval.com |
| 16 | claude-opus-4-6_unknownAnthropic | closed | effort unknown | 44.1% | 2 May 2026 | gbaeval.com |
| 17 | claude-opus-4-7_unknownAnthropic | closed | effort unknown | 43.8% | 2 May 2026 | gbaeval.com |
| 18 | muse-spark-1.1Meta AI | closed | default | 7.9% | 17 Jul 2026 | gbaeval.com |
| 19 | gemini-3.5-flash_unknownGoogle DeepMind | closed | effort unknown | 6.7% | 20 May 2026 | gbaeval.com |
| 20 | grok-build-0.1xAI | closed | default | 2.4% | 8 Jun 2026 | gbaeval.com |
| 21 | MiniMaxAI/MiniMax-M3Epoch: MiniMax-M3, MiniMax | open | default | 0.9% | 2 Jun 2026 | gbaeval.com |
| 22 | moonshotai/Kimi-K2.6Epoch: kimi-k2.6, Moonshot | open | default | 0.9% | 1 May 2026 | gbaeval.com |
| 23 | moonshotai/Kimi-K2.7-CodeEpoch: kimi-k2.7-code, Moonshot | open | default | 0.9% | 21 Jun 2026 | gbaeval.com |
| 24 | gemini-3.1-pro-previewGoogle DeepMind | closed | default | 0.8% | 3 May 2026 | gbaeval.com |
| 25 | qwen3.7-maxAlibaba | closed | default | 0.4% | 3 Jun 2026 | gbaeval.com |
| 26 | zai-org/GLM-5.2Epoch: glm-5.2_unknown, Z.ai (Zhipu AI) | open | effort unknown | 0.0% | 21 Jun 2026 | gbaeval.com |
| 27 | zai-org/GLM-5.1Epoch: glm-5.1, Z.ai (Zhipu AI) | open | default | 0.0% | not dated | gbaeval.com |
| 28 | MiniMaxAI/MiniMax-M2.7Epoch: MiniMax-M2.7, MiniMax | open | default | 0.0% | not dated | gbaeval.com |
Open-weight rows are shaded. Dates are when the result was run or evaluated, where the data says. Data from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Epoch AI. Method