Individual benchmark scores plotted by date.
| Organisation | Model | Reported | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|---|
| Claude Mythos Preview | 07 Apr 2026 | 92.10% | Terminal-Bench 2.1 updates; 4h timeout | Yes | Source | |
| GPT 5.5 | 23 Apr 2026 | 82.70% | - | Yes | Source | |
| GPT 5.3 Codex | 05 Feb 2026 | 77.30% | xhigh reasoning | Yes | Source | |
| GPT 5.4 | 05 Mar 2026 | 75.10% | - | Yes | Source | |
| GPT 5 Search API | 14 Oct 2025 | 75.10% | inferred family alias from gpt-5.4 (score=0.3050; benches=19) | Yes | Source | |
| GPT 5 Pro | 07 Aug 2025 | 75.10% | inferred family alias from gpt-5.4 (score=0.4083; benches=19) | Yes | Source | |
| Qwen 3.7 Max | 21 May 2026 | 69.70% | Harbor/Terminus-2 harness; avg of 5 runs | Yes | Source | |
| Claude Opus 4.7 | 16 Apr 2026 | 69.40% | - | Yes | Source | |
| GLM 5.1 | 07 Apr 2026 | 69% | Best self-reported (Claude Code) | Yes | Source | |
| Gemini 3.1 Pro Preview | 19 Feb 2026 | 68.50% | Terminus-2 Harness | Yes | - | |
| DeepSeek V4 Pro | 24 Apr 2026 | 67.90% | Acc; DeepSeek-V4-Pro Max | Yes | Source | |
| MiMo V2.5 | 22 Apr 2026 | 65.80% | Hugging Face evaluation result | Yes | Source | |
| Claude Opus 4.6 | 05 Feb 2026 | 65.40% | - | Yes | Source | |
| GPT 5.2 Codex | 18 Dec 2025 | 64% | - | Yes | Source | |
| Qwen 3.6 Plus | 01 Apr 2026 | 61.60% | - | Yes | Source | |
| GPT 5.4 Mini | 17 Mar 2026 | 60% | - | Yes | Source | |
| Claude Opus 4.5 | 24 Nov 2025 | 59.30% | Avg@5, 128k Thinking | Yes | Source | |
| Claude Sonnet 4.6 | 17 Feb 2026 | 59.10% | Default thinking (no thinking budget) | Yes | Source | |
| Muse Spark | 08 Apr 2026 | 59% | - | Yes | Source | |
| GPT 5.3 Codex Spark | 12 Feb 2026 | 58.40% | - | Yes | Source | |
| MiMo V2 Pro | 18 Mar 2026 | 57.10% | - | Yes | Source | |
| MiMo V2 TTS | 18 Mar 2026 | 57.10% | inferred modality/version alias from mimo-v2-pro | Yes | Source | |
| MiniMax M2.7 | 18 Mar 2026 | 57% | - | Yes | Source | |
| DeepSeek V4 Flash | 24 Apr 2026 | 56.90% | Acc; DeepSeek-V4-Flash Max | Yes | Source | |
| GLM 5 | 11 Feb 2026 | 56.20% | - | Yes | Source | |
| Seed 2.0 Pro | 14 Feb 2026 | 55.80% | Seed2 official benchmark table | Terminal Bench 2.0 1 | Yes | Source | |
| Gemini 3 Pro Preview | 18 Nov 2025 | 54.20% | - | Yes | Source | |
| Gemini 3 Pro Image Preview (Nano Banana Pro) | 20 Nov 2025 | 54.20% | inferred modality/version alias from gemini-3-pro-preview | Yes | Source | |
| GPT Audio 1.5 | 23 Feb 2026 | 52.80% | manual fallback alias from gpt-5.1-codex | Yes | Source | |
| GPT Realtime 1.5 | 23 Feb 2026 | 52.80% | manual fallback alias from gpt-5.1-codex | Yes | Source | |
| GPT 5.1 Codex | 13 Nov 2025 | 52.80% | - | Yes | Source | |
| Qwen 3.5 397B A17B | 16 Feb 2026 | 52.50% | - | Yes | Source | |
| Step 3.5 Flash | - | 51% | - | Yes | Source | |
| Kimi K2.5 | 27 Jan 2026 | 50.80% | - | Yes | Source | |
| Qwen 3.5 122B A10B | 24 Feb 2026 | 49.40% | - | Yes | Source | |
| Gemini 3 Flash Preview | 17 Dec 2025 | 47.60% | - | Yes | Source | |
| DeepSeek V3.2 Speciale | 01 Dec 2025 | 46.40% | - | Yes | Source | |
| GPT 5.4 Nano | 17 Mar 2026 | 46.30% | - | Yes | Source | |
| Seed 2.0 Lite | 14 Feb 2026 | 45% | Seed2 official benchmark table | Terminal Bench 2.0 1 | Yes | Source | |
| Qwen 3.5 27B | 24 Feb 2026 | 41.60% | - | Yes | Source | |
| Qwen 3.5 Flash | 23 Feb 2026 | 41.60% | inferred family alias from qwen3.5-27b (score=0.4147; benches=81) | Yes | Source | |
| GLM 4.7 | 22 Dec 2025 | 41% | - | Yes | Source | |
| Qwen 3.5 35B A3B | 24 Feb 2026 | 40.50% | - | Yes | Source | |
| MiMo V2 Flash | 16 Dec 2025 | 38.50% | - | Yes | Source | |
| Nemotron 3 Super | 11 Mar 2026 | 31% | - | Yes | Source |
