Atlas

Benchmarks

← All benchmarks

Market-Bench - Single-Stock Scheduled Execution (Best of 5)

Professional Work · 2025-12-13

The easy Market-Bench strategy asks a model to implement scheduled buying and selling of MSFT against randomized market-by-price L10 liquidity while correctly reserving consumed synthetic-book volume. This definition reports the lowest mean absolute error among five generated backtester runs.

Top models (lower is better)

ModelScore
GPT-5.1-Codex-Max0.0
Opus 4.57.2
Grok 47.2
DeepSeek-V3.27.3
Gemini 3 Pro Preview14.8
Sonnet 4.516.4
Qwen3-Max (2025-09-23)16.4
Mistral Large 3 675B Instruct 251223.0
Llama 3.1 Nemotron Ultra 253B V1112
GPT-5.2153
Command A153
Nova Premier153
Llama 4 Maverick170
Loading Atlas data…