Reasoning
APEX-Agents Leaderboard
APEX-Agents (AI Productivity Index) is Mercor's evaluation of AI agents on long-horizon, multi-application professional-services work — investment-banking, management-consulting, and corporate-law style tasks — graded against expert-produced solutions. The score is Pass@1, the fraction solved correctly on the first attempt.
Source: epoch20 open models ranked+40 proprietaryData through Oct 2026
Open models ranked on APEX-Agents
# shows rank among open models / rank overall (including proprietary).
| # | Model | Score | VRAM @ Q4 |
|---|---|---|---|
| 1 / 4 | DeepSeek V4.1 Flash 1017 · 763.2B | 69.8% | ~458 GB |
| 2 / 19 | GLM 5.3 · 753.3B | 56.6% | ~456 GB |
| 3 / 25 | MiMo V2.6 Pro RL · 1020B | 54.4% | ~613 GB |
| 4 / 27 | GLM 5.3 Flash · 321.3B | 52.8% | ~195 GB |
| 5 / 30 | Mistral Large 4 1T A52B · 1000B | 51.6% | ~604 GB |
| 6 / 34 | Beam 1.0 501B A23B · 501.3B | 48.8% | ~301 GB |
| 7 / 35 | Kimi K3 · 2779.9B | 47.8% | ~1674 GB |
| 8 / 37 | Qwen3.8 27B · 27.8B | 47.5% | ~17 GB |
| 9 / 38 | DeepSeek V4 Pro 0813 · 1650.5B | 47.3% | ~991 GB |
| 10 / 42 | GLM 5.2 · 753.3B | 45.2% | ~456 GB |
| 11 / 47 | GLM 5.1 · 753.9B | 40.9% | ~457 GB |
| 12 / 48 | DeepSeek V4.1 Flash · 763.2B | 39.5% | ~458 GB |
| 13 / 49 | MiniMax M3 · 427.0B | 37.7% | ~257 GB |
| 14 / 50 | Kimi K2.7 Code · 1026.9B | 37.6% | ~620 GB |
| 15 / 52 | Inkling · 952.4B | 33.8% | ~572 GB |
| 16 / 53 | Muse Spark Open 816B A42B · 822.0B | 33.4% | ~543 GB |
| 17 / 57 | Qwen3.5 397B A17B · 403.4B | 24.9% | ~242 GB |
| 18 / 58 | NVIDIA Nemotron 3 Ultra 550B A55B BF16 · 560.5B | 22.7% | ~370 GB |
| 19 / 59 | DeepSeek V3.2 · 685.4B | 21.3% | ~415 GB |
| 20 / 60 | GPT OSS 120B · 116.8B | 4.4% | ~71 GB |
Score vs model size
Which models give the most quality for their size — the ones worth running locally.
- Qwen3.8 27B, 28B, score 47.5% — on the efficiency frontier (best score at its size or smaller).
- GLM 5.3 Flash, 321B, score 52.8% — on the efficiency frontier (best score at its size or smaller).
- GLM 5.3, 753B, score 56.6% — on the efficiency frontier (best score at its size or smaller).
- DeepSeek V4.1 Flash 1017, 763B, score 69.8% — on the efficiency frontier (best score at its size or smaller).
APEX-Agents: frequently asked questions
- What is the best open LLM on APEX-Agents?
- DeepSeek V4.1 Flash 1017 is the top open model on APEX-Agents, scoring 69.8%. Among all models tested — including proprietary ones — it ranks #4. The top model overall is Gemini 4 Argon (unspecified) (Google DeepMind) at 82.4%.
- What's the best APEX-Agents model you can run on a 24 GB GPU?
- Qwen3.8 27B is the highest-scoring open model that fits in 24 GB at 4-bit quantization (~17 GB), scoring 47.5% on APEX-Agents.
- Can open models match proprietary models on APEX-Agents?
- Not quite on APEX-Agents: the strongest proprietary model (Gemini 4 Argon (unspecified)) scores 82.4%, ahead of the best open model (DeepSeek V4.1 Flash 1017) at 69.8% — but you can run the open one yourself.
Scores aggregated from epoch. llmrun does not run this benchmark — see the source for methodology, or the about benchmarks for what it measures.