Live LLM benchmark: new models and ranking changes
What moved on Code Arena, Text Arena and LiveBench. Every model that entered a leaderboard, every new #1 and every overtake inside a top 10, read from the leaderboards twice a day.
Last crawl per source
The crawl is scheduled for 06:00 and 18:00 UTC and often finishes hours later. The table shows when each run fetched. The source date is the leaderboard's own update, which can be older than the crawl.
| Source | Fetched | Source date | Models |
|---|---|---|---|
| Text Arena | 8 Oct 2026 | 414 | |
| Code Arena | 8 Oct 2026 | 142 | |
| LiveBench | release 2026-11-06 | 36 |
Twice a day, scheduled 06:00 and 18:00 UTC
New #1s
Every change at the top of a leaderboard in the last 90 days.
- Gemini 4 Carbon (Google) took #1 on LiveBench Coding
- Gemini 4 Carbon (Google) took #1 on LiveBench Data Analysis
- Claude Fable 5.2 (xhigh) (Anthropic) took #1 on LiveBench Overall
- Claude Fable 5.1 (Anthropic) took #1 on LiveBench Coding (from #2)
- Gemini 4 Argon High (Google) took #1 on Text Arena
- Claude Opus 5.5 High (Anthropic) took #1 on Text Arena
- Claude Opus 5.5 Max (Anthropic) took #1 on Code Arena
- Claude Fable 5 High (Anthropic) took #1 on Text Arena
- GPT-6 Astra Max (OpenAI) took #1 on Code Arena
- Claude Fable 5.1 Max (Anthropic) took #1 on Code Arena
Ranking changes, day by day
The last 30 days. A climb counts when a model overtakes another one inside a top 10 and its rank changes. Models that fell behind are counted, not listed.
- Qwen 4 Max (Alibaba): LiveBench Data Analysis entered at #3, LiveBench Reasoning entered at #5, LiveBench Overall entered at #11, LiveBench Coding entered at #22
- Gemini 4 Hydrogen (Google): LiveBench Data Analysis entered at #3, LiveBench Overall entered at #9, LiveBench Coding entered at #13, LiveBench Reasoning entered at #18
- Gemini 4 Carbon (Google): took #1 on LiveBench Coding, took #1 on LiveBench Data Analysis, LiveBench Overall entered at #4, LiveBench Reasoning entered at #10
- Muse Spark 1.3 (Meta): LiveBench Coding climbed from #8 to #7
- 3 models moved down
- Claude Fable 5.2 (xhigh) (Anthropic): took #1 on LiveBench Overall, LiveBench Data Analysis entered at #6, LiveBench Coding entered at #11, LiveBench Reasoning entered at #15
- GPT-6.1 Terra (OpenAI): LiveBench Coding climbed from #7 to #2, LiveBench Overall climbed from #7 to #6
- GPT-6 Astra (OpenAI): LiveBench Coding climbed from #8 to #3
- Gemini 4 Argon (Google): LiveBench Coding climbed from #11 to #5
- GPT-5.6 Terra (OpenAI): LiveBench Coding climbed from #15 to #6
- Muse Spark 1.3 (Meta): LiveBench Coding climbed from #10 to #7
- 9 models moved down
- Mistral Large 4 1029 (Mistral AI): LiveBench Coding entered at #5, LiveBench Data Analysis entered at #15, LiveBench Reasoning entered at #17, LiveBench Overall entered at #19
- GPT-6.1 Terra (OpenAI): LiveBench Data Analysis entered at #4, LiveBench Coding entered at #6, LiveBench Overall entered at #7, LiveBench Reasoning entered at #15
- Claude Fable 5.1 (Anthropic): took #1 on LiveBench Coding
- GPT-6 Astra (OpenAI): LiveBench Overall climbed from #3 to #2
- Claude Sonnet 5.5 (Anthropic): LiveBench Overall climbed from #7 to #5
- Muse Wonder 1.0 (Meta): LiveBench Overall climbed from #11 to #7
- Grok 4.7 (xAI): LiveBench Overall climbed from #21 to #10
- 2 models moved down
- Kimi K3.1 Pro (Moonshot AI): LiveBench Reasoning entered at #10, LiveBench Overall entered at #14, LiveBench Data Analysis entered at #16, LiveBench Coding entered at #29
- DeepSeek V4 Flash 0731 (DeepSeek): LiveBench Coding entered at #34, LiveBench Data Analysis entered at #38, LiveBench Overall entered at #39, LiveBench Reasoning entered at #49
- Smaug Nemotron 3.5 NVFP4 (Abacus.AI): LiveBench Coding entered at #38, LiveBench Data Analysis entered at #42, LiveBench Overall entered at #44, LiveBench Reasoning entered at #44
- Nemotron 3.5 Super VL (NVFP4) (NVIDIA): LiveBench Data Analysis entered at #41, LiveBench Coding entered at #44, LiveBench Overall entered at #47, LiveBench Reasoning entered at #48
- MiniMax M3.1 Flash (MiniMax): LiveBench Overall entered at #21, LiveBench Coding entered at #25, LiveBench Reasoning entered at #34, LiveBench Data Analysis entered at #35
- Claude Haiku 5.5 (Anthropic): LiveBench Coding climbed from #22 to #10
- 1 model moved down
Oldest recorded change: 22 May 2026.
Need a model picked for your use case?
Benchmarks rank models on other people's tasks. For a business workflow I test the shortlist on your own data, costs and privacy requirements, and recommend one.