U2 Flash added, ERNIE-Light-DT-1.0 added (11 Nov 2026)
U2 Flash and ERNIE-Light-DT-1.0 join the board. 2 new models, 7 top-10 rank changes, 5 price changes, 203 new benchmark results. Sources linked.
BenchLeader refreshed 1693 models from 28 of 29 data feeds (benchmark publishers, aggregators and price lists). 2 new models, 7 top-10 rank changes, 5 price changes, 203 new or updated benchmark results across 28 data feeds.
New models
- U2 Flash (Unisound) appeared with 1 benchmark result. (details)
- ERNIE-Light-DT-1.0 (Baidu) is now listed at $1.83/M blended; no independent results yet. (details)
Movement in the top 10
- Claude Fable 5.2 (xhigh) moved from #4 to #2 in the BenchLeader Index. (details)
- Gemini 4 Carbon (high) moved from #5 to #3 in the BenchLeader Index. (details)
- Claude Opus 5.5 (max) moved from #3 to #4 in the BenchLeader Index. (details)
- Claude Fable 5.2 (high) moved from #2 to #5 in the BenchLeader Index. (details)
- GPT-6 Astra moved from #6 to #8 in the BenchLeader Index. (details)
- GPT-6.1 Sol (max) moved from #8 to #10 in the BenchLeader Index. (details)
- Claude Fable 5 moved from #10 to #13 in the BenchLeader Index. (details)
Price changes
- Llama 4 Maverick output price moved from $0.92/M to $0.97/M (5% up). (details)
- Qwen2.5 7B Instruct input price moved from $0.10/M to $0.050/M (50% down). (details)
- Qwen2.5 7B Instruct output price moved from $0.20/M to $0.050/M (75% down). (details)
- Nemotron 3 Nano Omni 30B A3B (thinking) input price moved from $0.50/M to $0.20/M (60% down). (details)
- Nemotron 3 Nano Omni 30B A3B (thinking) output price moved from $0.90/M to $0.80/M (11% down). (details)
New benchmark results
- MirrorCode: 29 new results, including Claude Opus 5.5, GPT-6 Astra, Claude Fable 5.1… (board)
- SimpleBench: 22 new results, including Claude Fable 5.2, GPT-6 Astra, Claude Fable 5.1… (board)
- GeoBench: 17 new results, including Claude Fable 5, GPT-5.6 Sol, Claude Fable 5.1… (board)
- CursorBench: 8 new results, including Claude Fable 5.2, Claude Fable 5.2, Gemini 4 Carbon… (board)
- SciCode: 5 new results, including Claude Fable 5.2, Claude Fable 5.2, Claude Fable 5.2… (board)
- LMCA: 5 new results, including Claude Fable 5.2, Claude Fable 5.2, Claude Fable 5.2… (board)
- DTBench: 5 new results, including Claude Fable 5.2, Claude Fable 5.2, Claude Fable 5.2… (board)
- ALE-Bench: 5 new results, including Claude Fable 5.2, Qwen3.5 397B A17B, Qwen3.5-35B-A3B… (board)
- ARC-AGI-1: 3 new results, including Gemini 4 Carbon, Gemini 4 Carbon, Gemini 4 Carbon (board)
- ARC-AGI-2: 3 new results, including Gemini 4 Carbon, Gemini 4 Carbon, Gemini 4 Carbon (board)
- ARC-AGI-3: 3 new results, including Gemini 4 Carbon, Gemini 4 Carbon, Gemini 4 Carbon (board)
- APEX-Agents: 2 new results, including Claude Fable 5.2, Muse Ember (board)
Score revisions
- Claude Opus 5.5 (max) on Coding-agent share: 57.1% → 56.6%. (details)
- Claude Fable 5.2 on Design Arena Website: 1298 → 1320. (details)
- Claude Fable 5.2 on Design Arena Agentic: 1346 → 1362. (details)
- GPT-6 Astra (max) on Design Arena Agentic: 1391 → 1385. (details)
- GPT-6.1 Sol (max) on Design Arena Agentic: 1423 → 1398. (details)
- Claude Opus 5.5 (xhigh) on Coding-agent share: 57.1% → 56.6%. (details)
- Claude Opus 5.5 (high) on Design Arena Agentic: 1459 → 1440. (details)
- Claude Opus 5.5 (high) on Coding-agent share: 57.1% → 56.6%. (details)
- Gemini 4 Argon on Design Arena Agentic: 1277 → 1264. (details)
- Claude Opus 5.5 on Coding-agent share: 57.1% → 56.6%. (details)
- Kimi K3 (max) on GameBench 2 by reasoning setting: 43.8 → 44.8. (details)
- Muse Wonder (max) on Design Arena Agentic: 1164 → 1156. (details)
- Claude Opus 5.5 (medium) on Coding-agent share: 57.1% → 56.6%. (details)
- GPT-5.6 Sol (medium) on GameBench 2 by reasoning setting: 65.8 → 65.2. (details)
- Gemini 3.6 Flash (high) on GameBench 2 by reasoning setting: 62.8 → 62.1. (details)
- Grok 4.6 (medium) on GameBench 2 by reasoning setting: 68.6 → 69.4. (details)
- Claude Opus 5.5 (low) on Coding-agent share: 57.1% → 56.6%. (details)
- DeepSeek V4.1 Flash 1017 (max) on OpenRouter token share: 26.6% → 26%. (details)
- GPT-5 (medium) on GameBench 2 by reasoning setting: 37.2 → 36.6. (details)
- GPT-5.4 on GameBench 2 Overall: 53.4 → 54.1. (details)
- …and 40 more.
Speed changes
- GPT-5.6 Sol time to first token changed from 4.52 s to 2.24 s. (details)
- GPT-5.5 time to first token changed from 3.70 s to 1.96 s. (details)
- GPT-6.1 Sol output speed changed from 34 tok/s to 46 tok/s. (details)
- Kimi K3 time to first token changed from 1.48 s to 0.88 s. (details)
- Kimi K2.6 time to first token changed from 0.54 s to 0.77 s. (details)
- Kimi K3 output speed changed from 59 tok/s to 82 tok/s. (details)
- Inkling 2 Small time to first answer changed from 32.28 s to 46.89 s. (details)
- Inkling 2 Small thinking time changed from 29.97 s to 44.07 s. (details)
- GPT-5.2-Codex output speed changed from 20 tok/s to 45 tok/s. (details)
- GPT-5.2-Codex time to first token changed from 2.99 s to 1.82 s. (details)
- GPT-5 nano output speed changed from 41 tok/s to 90 tok/s. (details)
- GPT-5 nano time to first token changed from 1.58 s to 2.43 s. (details)
- GLM 5V Turbo output speed changed from 36 tok/s to 50 tok/s. (details)
- DeepSeek V3.1 Terminus output speed changed from 3 tok/s to 20 tok/s. (details)
- DeepSeek V3.1 Terminus time to first token changed from 21.54 s to 1.89 s. (details)
- …and 53 more.
Today's top five
- Claude Fable 5.2 (max) — 72.7
- Gemini 4 Carbon (high) — 68.9
- Claude Opus 5.5 (max) — 68.4
- GPT-6 Astra (max) — 67.3
- Claude Fable 5.1 (max) — 66.7
More: New AI model releases this week, and the full LLM leaderboard 2026.
This digest is generated automatically from the day's data changes. Every line links to the page where you can check the numbers and their source.