← Whole ranking · mysummit.school AI model benchmark
Compare two AI models
Two models from the benchmark side by side: the overall level, access from Russia, the API price and the 8 task categories.
An example pair – choose your own models.
On the overall score they are on one level. Choose by access, API price and the category you need. From Russia without a VPN it is simpler to start with Kimi K3.
Where they differ
A – GPT-5.6 Sol B – Kimi K3
-
Information search
higher for GPT-5.6 Sol in this run
-
Data analysis
on one level
-
Decisions under uncertainty
on one level
-
Feedback analysis
on one level
-
Corporate requirements and risks
on one level
-
Automation
on one level
-
People management
on one level
-
Russia and Kazakhstan specifics
on one level
Two answers to one task
This pair has no category with a gap of 0.5 points or more – the answers would not show a difference worth reading.
How to read the comparison
- Scores are from one run per model, judged by two AI judges.
- A difference under 0.3 points is treated as one level – a heuristic for one run, not a proof.
- Data as of July 2026; versions may have changed.
Every model’s category scores
| Model | Information search | Data analysis | Decisions under uncertainty | Feedback analysis | Corporate requirements and risks | Automation | People management | Russia and Kazakhstan specifics |
|---|---|---|---|---|---|---|---|---|
| Alice AI LLM | 5.7 | 6.2 | 6.5 | 6.9 | 6.2 | 5.5 | 6.8 | 4.7 |
| Alice AI LLM (v3) | 6.1 | 6.4 | 6.9 | 6.9 | 6.5 | 5.8 | 6.8 | 5.1 |
| Claude Fable 5 | 8.1 | 8.6 | 9.0 | 9.3 | 8.7 | 8.4 | 9.2 | 8.8 |
| Claude Haiku 4.5 | 6.6 | 7.4 | 8.1 | 8.1 | 6.9 | 7.6 | 7.8 | 5.8 |
| Claude Opus 4.6 | 8.3 | 8.6 | 8.9 | 9.2 | 8.7 | 8.7 | 9.1 | 8.7 |
| Claude Opus 4.7 | 7.7 | 8.2 | 8.5 | 8.9 | 8.2 | 8.1 | 8.7 | 8.3 |
| Claude Opus 4.8 | 7.4 | 8.1 | 8.4 | 8.9 | 8.0 | 8.1 | 8.7 | 8.3 |
| Claude Sonnet 4.6 | 7.7 | 8.0 | 8.5 | 8.9 | 8.0 | 8.1 | 8.6 | 7.7 |
| Claude Sonnet 5 | 7.5 | 7.8 | 8.1 | 8.4 | 7.8 | 8.3 | 8.6 | 7.7 |
| DeepSeek V4 Flash | 7.2 | 7.5 | 8.0 | 8.2 | 7.1 | 6.9 | 7.7 | 7.0 |
| DeepSeek V4 Flash (Yandex Cloud) | 7.4 | 6.8 | 7.9 | 8.2 | 7.3 | 6.9 | 7.9 | 6.2 |
| DeepSeek V4 Pro | 7.5 | 7.6 | 7.8 | 8.4 | 7.6 | 7.5 | 8.3 | 7.4 |
| Gemini 3 Flash | 6.8 | 6.9 | 7.2 | 7.8 | 6.7 | 6.0 | 7.2 | 6.1 |
| Gemini 3.1 Flash Lite | 6.4 | 6.3 | 6.6 | 7.2 | 6.1 | 5.5 | 6.6 | 5.2 |
| Gemini 3.1 Pro | 7.1 | 7.5 | 7.7 | 8.0 | 7.3 | 6.5 | 8.0 | 7.3 |
| Gemini 3.5 Flash | 7.3 | 7.4 | 7.8 | 8.2 | 7.3 | 7.2 | 8.2 | 7.8 |
| Gemma 4 26B MoE | 6.5 | 6.7 | 7.0 | 7.6 | 6.5 | 6.0 | 7.1 | 5.6 |
| Gemma 4 31B | 6.6 | 7.0 | 7.4 | 7.6 | 6.7 | 6.0 | 7.4 | 6.2 |
| GigaChat 2 Max | 3.9 | 4.7 | 4.4 | 4.3 | 4.3 | 4.0 | 4.4 | 3.6 |
| GigaChat 3.5 Ultra | 7.0 | 7.5 | 8.1 | 7.4 | 7.0 | 6.3 | 6.3 | 5.3 |
| GigaChat Ultra | 4.7 | 4.7 | 5.4 | 5.9 | 4.3 | 4.5 | 5.2 | 4.0 |
| GLM 5.1 | 7.2 | 7.7 | 7.9 | 7.7 | 6.8 | 7.0 | 8.1 | 6.6 |
| GLM 5.2 | 7.1 | 7.4 | 7.6 | 8.2 | 7.0 | 6.7 | 7.9 | 6.7 |
| GPT-5.4 Mini | 7.3 | 7.5 | 8.0 | 8.3 | 7.7 | 7.5 | 8.3 | 6.5 |
| GPT-5.4 Nano | 6.9 | 7.5 | 7.9 | 7.8 | 7.2 | 7.2 | 8.0 | 5.6 |
| GPT-5.5 | 8.4 | 8.4 | 8.9 | 9.1 | 8.5 | 8.5 | 9.1 | 8.4 |
| GPT-5.6 Luna | 8.4 | 8.4 | 8.8 | 8.9 | 8.5 | 8.7 | 9.1 | 7.9 |
| GPT-5.6 Sol | 8.9 | 8.8 | 9.1 | 9.3 | 9.0 | 8.9 | 9.4 | 8.8 |
| GPT-5.6 Terra | 8.7 | 8.8 | 9.1 | 9.3 | 8.9 | 8.9 | 9.3 | 8.6 |
| Grok 4.3 | 7.3 | 7.3 | 7.8 | 7.5 | 7.2 | 6.6 | 7.3 | 6.0 |
| Grok 4.5 | 7.4 | 8.6 | 8.6 | 8.9 | 8.2 | 8.6 | 9.1 | 7.4 |
| Hunyuan 3 (HY3) | 6.6 | 7.4 | 7.5 | 7.5 | 6.6 | 5.8 | 7.5 | 5.8 |
| Kimi K2.6 | 8.2 | 8.4 | 8.5 | 8.9 | 7.9 | 8.3 | 8.8 | 7.2 |
| Kimi K2.7 Code | 7.9 | 8.1 | 8.3 | 8.7 | 7.9 | 8.1 | 8.6 | 6.8 |
| Kimi K3 | 8.5 | 9.0 | 9.1 | 9.3 | 8.8 | 8.9 | 9.3 | 8.6 |
| MiMo v2.5 | 7.5 | 8.1 | 8.1 | 8.5 | 7.7 | 8.0 | 8.3 | 6.4 |
| MiMo v2.5 Pro | 7.8 | 8.4 | 8.6 | 9.0 | 8.1 | 8.5 | 8.8 | 7.9 |
| MiniMax M2.7 | 7.2 | 7.5 | 8.3 | 8.2 | 7.5 | 7.6 | 8.2 | 6.7 |
| MiniMax M3 | 7.8 | 8.6 | 8.7 | 9.0 | 8.4 | 8.4 | 8.7 | 7.5 |
| Mistral Medium 3.5 | 6.5 | 6.7 | 7.5 | 7.4 | 6.9 | 6.8 | 7.4 | 5.7 |
| Nemotron 3 Ultra 550B | 8.2 | 8.0 | 8.8 | 9.0 | 8.4 | 8.7 | 8.7 | 6.5 |
| Qwen 3.6 27B | 7.3 | 7.2 | 8.0 | 8.8 | 7.4 | 7.6 | 8.2 | 6.3 |
| Qwen 3.6 Flash | 7.6 | 7.8 | 8.2 | 8.3 | 7.7 | 7.2 | 8.2 | 5.7 |
| Qwen 3.6 Plus | 7.9 | 8.1 | 8.3 | 8.8 | 7.5 | 7.5 | 8.3 | 7.1 |
| Qwen 3.7 Max | 7.7 | 7.8 | 7.9 | 8.5 | 7.7 | 7.3 | 8.2 | 7.2 |
| Qwen 3.7 Plus | 7.5 | 7.4 | 7.8 | 8.1 | 7.4 | 7.0 | 8.0 | 6.6 |
| Step 3.7 Flash | 6.7 | 7.1 | 7.7 | 7.4 | 6.2 | 7.0 | 7.5 | 5.4 |
A model from the top group will not fix a vague task
The free first lessons show how to set context and the shape of the answer, and how to check the result on your own work task. The course is in Russian.
Open the free first lessons