Recap · 2025-09
AI leaderboard changes September 2025
As of Sep 29, 2025, 10 confirmed crown changes are recorded for September 2025.
Crowns
Confirmed crown changes
- Sep 29, 2025. Math: O3 and Gemini 2.5 Pro Preview 06.05 and GPT-5 and GPT-5 Mini are co-SOTA. Crown change confirmed by two consecutive snapshots or two independent primary sources. O3Gemini 2.5 Pro Preview 06.05GPT-5GPT-5 Mini · Math
- Sep 18, 2025. Image edit: Qwen Image Edit takes the crown. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen Image Edit · Image edit
- Sep 18, 2025. Long context: Qwen3 235B A22b Instruct 2507 takes the crown. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen3 235B A22b Instruct 2507 · Long context
- Sep 18, 2025. General LLM: Qwen3 235B A22b Instruct 2507 takes the crown. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen3 235B A22b Instruct 2507 · General LLM
- Sep 17, 2025. Long context: Gemini 2.5 Pro takes the crown. Crown change confirmed by two consecutive snapshots or two independent primary sources. Gemini 2.5 Pro · Long context
- Sep 17, 2025. General LLM: Gemini 2.5 Pro takes the crown. Crown change confirmed by two consecutive snapshots or two independent primary sources. Gemini 2.5 Pro · General LLM
- Sep 8, 2025. Reasoning: Claude Opus 4.1 and Gemini 2.5 Pro and GPT-5 are co-SOTA. Crown change confirmed by two consecutive snapshots or two independent primary sources. Claude Opus 4.1Gemini 2.5 ProGPT-5 · Reasoning
- Sep 4, 2025. Embeddings: Qwen Qwen3 Embedding 8B and Qwen Qwen3 Embedding 4B and Google Gemini Embedding 001 are co-SOTA. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen Qwen3 Embedding 8BQwen Qwen3 Embedding 4BGoogle Gemini Embedding 001 · Embeddings
- Sep 4, 2025. Embeddings: Qwen Qwen3 Embedding 8B and Qwen Qwen3 Embedding 4B are co-SOTA. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen Qwen3 Embedding 8BQwen Qwen3 Embedding 4B · Embeddings
- Sep 4, 2025. Embeddings: Qwen Qwen3 Embedding 8B and Qwen Qwen3 Embedding 4B are co-SOTA. Crown change confirmed by two consecutive snapshots or two independent primary sources. Qwen Qwen3 Embedding 8BQwen Qwen3 Embedding 4B · Embeddings
Top 3
Models that entered the top 3
- Claude Opus 4.1 at #1 in Reasoning
- Gemini 2.5 Pro at #1 in General LLM
- Gemini 2.5 Pro at #1 in Reasoning
- Gemini 2.5 Pro at #1 in Long context
- Gemini 2.5 Pro Preview 06.05 at #1 in Math
- Google Gemini Embedding 001 at #1 in Embeddings
- GPT-5 at #1 in Reasoning
- GPT-5 at #1 in Math
- GPT-5 Mini at #1 in Math
- O3 at #1 in Math
- Qwen Image Edit at #1 in Image edit
- Qwen Qwen3 Embedding 4B at #1 in Embeddings
- Qwen Qwen3 Embedding 8B at #1 in Embeddings
Benchmarks
Benchmarks added or saturated
Benchmark additions for this month are still empty.