Step 5 Preview + 12 new AI models: 14 Sep–20 Sep

The week of 14 Sep 2026 to 20 Sep 2026, from 7 daily updates.

Three things that happened

• Biggest riser: GLM 4.5V, up 8.0 points to 51.5 on the BenchLeader Index as new results landed.
• 13 new models arrived: Step 5 Preview, GLM 5.3 FlashX, union-alpha, Jamba Large, Jamba Mini and more.
• Busiest benchmark: CritPt with 436 new results; 6954 new results were published in all.

Index movers

• Nemotron 3 Ultra 550B A55B-8.4
• GLM 4.5V+8.0
• Grok 4.3+7.6
• GPT-5.4 Mini+6.7
• Qwen3 14B+6.5
• Qwen3 30B A3B+6.0
• Grok 4.20+5.7
• Qwen3 32B+5.3

| Model | Provider | 14 Sep 2026 | 20 Sep 2026 | Change | |---|---|---:|---:|---:| | Nemotron 3 Ultra 550B A55B | NVIDIA | 60.6 | 52.2 | -8.4 | | GLM 4.5V | Zhipu AI | 43.5 | 51.5 | +8.0 | | Grok 4.3 | SpaceXAI | 50.7 | 58.3 | +7.6 | | GPT-5.4 Mini | OpenAI | 50.2 | 56.9 | +6.7 | | Qwen3 14B | Alibaba | 41.2 | 47.7 | +6.5 | | Qwen3 30B A3B | Alibaba | 42.9 | 48.9 | +6.0 | | Grok 4.20 | SpaceXAI | 54.3 | 60.0 | +5.7 | | Qwen3 32B | Alibaba | 46.8 | 52.1 | +5.3 | | GPT-5.2-Codex | OpenAI | 61.3 | 56.3 | -5.0 | | GPT-5.6 Terra | OpenAI | 58.7 | 63.4 | +4.7 |

New models

• Step 5 Preview (StepFun) appeared with 5 benchmark results, entering the BenchLeader Index at #39. (details)
• GLM 5.3 FlashX (Zhipu AI) is now listed at $0.59/M blended; no independent results yet.
• Union Alpha (Unknown) appeared with 1 benchmark result.
• Jamba Large (AI21 Labs) is now listed at $3.50/M blended; no independent results yet.
• Jamba Mini (AI21 Labs) is now listed at $0.25/M blended; no independent results yet.
• Qwen3 8 (Alibaba) appeared with 4 benchmark results. (details)
• Solar Pro 4 (Upstage) appeared with 1 benchmark result. (details)
• Palmyra X4 (Writer) is now listed at $4.38/M blended; no independent results yet. (details)
• Mercury 2.5 (high) (Inception) appeared with 3 benchmark results. (details)
• K2 Horizon 0.9B (MBZUAI) appeared with 3 benchmark results.
• K2 Horizon 3.7B (MBZUAI) appeared with 3 benchmark results. (details)
• K2 Horizon 7B (MBZUAI) appeared with 3 benchmark results. (details)
• K2 Horizon MoVA 36B A4B (MBZUAI) appeared with 3 benchmark results. (details)

Price changes

• Nemotron Nano 12B v2 VL input price moved from $0.20/M to $0.00/M (100% down).
• Qwen-VL OCR input price moved from $0.72/M to $0.04/M (94% down).
• Deepseek v4 Flash Vision input price moved from $0.44/M to $0.15/M (66% down).
• Qwen3 Coder Plus output price moved from $5.00/M to $2.30/M (54% down).
• Qwen3.5 Plus (thinking) output price moved from $3.44/M to $1.72/M (50% down).
• Qwen3.5 Flash output price moved from $1.72/M to $1.03/M (40% down).
• Qwen3.5-9B (no reasoning) input price moved from $0.14/M to $0.10/M (29% down).
• MiniMax-M1 input price moved from $0.55/M to $0.40/M (27% down).
• Muse Glimmer 30B output price moved from $1.50/M to $1.10/M (27% down).
• Muse Glimmer output price moved from $1.50/M to $1.20/M (20% down).
• Qwen3 Coder Next output price moved from $1.50/M to $1.20/M (20% down).
• Qwen3.8 27B (xhigh) input price moved from $0.50/M to $0.40/M (20% down).
• gpt-oss-20b (high) input price moved from $0.06/M to $0.05/M (17% down).
• Llama 4 Maverick input price moved from $0.26/M to $0.24/M (8% down).
• Gemma 4 26B A4B input price moved from $0.13/M to $0.12/M (8% down).

Top five as of 13 Nov 2026

1. Claude Fable 5.2 — 72.7
2. Gemini 4 Carbon — 69.5
3. Claude Opus 5.5 — 68.4
4. GPT-6 Astra — 67.1
5. Claude Fable 5.1 — 66.6

More: New AI model releases this week, and the full LLM leaderboard 2026.

Generated from the week's daily updates; every line links to the page where the numbers and their sources can be checked.

The week in dated, citable sentences: state of the leaderboard, week 38.

Head to head

• Claude Fable 5.2 vs Gemini 4 Carbon
• Claude Opus 5.5 vs Step 5 Preview
• Claude Fable 5.2 vs Claude Opus 5.5