ModelsCompareBest forBenchmarksStatusPricingAPI

What is OTIS Mock AIME 2024-2025?

45 competition-style problems written by students of the Olympiad Training for Individual Study (OTIS) program, with integer answers from 0 to 999. Epoch AI runs every model itself under one setup, so scores are directly comparable.

OTIS Mock AIME 2024-2025 scores by model

1
GPT-6 AstraOpenAI100% ↗
2
GPT-5.6 SolOpenAI100% ↗
3
Claude Fable 5.1Anthropic100% ↗
4
GPT-5.6 TerraOpenAI99.7% ↗
5
Qwen3.8-MaxQwen99.4% ↗
6
Grok 4.6xAI99.2% ↗
7
Claude Opus 5Anthropic98.9% ↗
8
Gemini 3.8 FlashGoogle98.9% ↗
9
GPT-5.6 LunaOpenAI98.3% ↗
10
Claude Opus 4.8Anthropic98.3% ↗
11
GPT-5.4OpenAI97.8% ↗
12
Claude Opus 4.7Anthropic97.8% ↗
13
Grok 4.5xAI97.8% ↗
14
Kimi K3Moonshot AI97.2% ↗
15
Gemini 3.7 FlashGoogle97.2% ↗
16
DeepSeek V4 ProDeepSeek96.7% ↗
17
GPT-5.2OpenAI96.1% ↗
18
Kimi K2.6Moonshot AI96.1% ↗
19
Gemini 3.1 Pro PreviewGoogle95.6% ↗
20
Kimi K2.7 CodeMoonshot AI95.6% ↗
21
Gemini 3.5 FlashGoogle95.6% ↗
22
Gemini 3 Flash PreviewGoogle95.6% ↗
23
Claude Sonnet 5Anthropic94.7% ↗
24
Claude Opus 4.6Anthropic94.4% ↗
25
Gemini 3.6 FlashGoogle94.2% ↗
26
GLM-5.3-FlashZ.ai93.9% ↗
27
Grok 4.3xAI93.3% ↗
28
Qwen3.7-PlusQwen93.3% ↗
29
Qwen3.6-PlusQwen93.3% ↗
30
GLM-5.1Z.ai93.3% ↗
31
Grok 4.20 ReasoningxAI92.2% ↗
32
GPT-5OpenAI91.4% ↗
33
GLM-5.3Z.ai91.1% ↗
34
Qwen3.6 27BQwen91.1% ↗
35
GPT-5.4 miniOpenAI88.9% ↗
36
gpt-oss-120bOpenAI88.9% ↗
37
Qwen3.5 397B-A17BQwen88.9% ↗
38
GPT-5.1OpenAI88.6% ↗
39
GPT-5.4 nanoOpenAI87.8% ↗
40
GPT-5 miniOpenAI86.7% ↗