GLM-4.7-FlashX price cut, GLM 5.3 price cut (26 Oct 2026)

BenchLeader refreshed 1616 models from 24 of 24 data feeds (benchmark publishers, aggregators and price lists). 136 price changes, 93 new or updated benchmark results across 24 data feeds.

Price changes

• Gemini 4 Argon (high) cost per task moved from $1.99/M to $1.62/M (19% down). (details)
• Muse Wonder (max) cost per task moved from $4.54/M to $8.63/M (90% up). (details)
• Muse Spark 1.3 (max) cost per task moved from $1.60/M to $3.69/M (130% up). (details)
• GPT-5.5 (xhigh) cost per task moved from $2.63/M to $2.10/M (20% down). (details)
• GPT-5.5 (high) cost per task moved from $1.54/M to $1.40/M (9% down). (details)
• Muse Spark 1.3 (xhigh) cost per task moved from $1.37/M to $2.15/M (57% up). (details)
• Gemini 3.8 Flash (high) cost per task moved from $1.24/M to $3.49/M (180% up). (details)
• Grok 4.7 (high) cost per task moved from $2.73/M to $2.36/M (13% down). (details)
• GLM 5.3 (max) cached input price moved from $0.30/M to $0.26/M (13% down). (details)
• Muse Spark 1.2 (xhigh) cost per task moved from $0.97/M to $1.29/M (32% up). (details)
• Gemini 3.5 Flash (high) cost per task moved from $3.69/M to $2.98/M (19% down). (details)
• GPT-5.5 (medium) cost per task moved from $0.90/M to $0.77/M (15% down). (details)
• GLM 5.2 (max) cost per task moved from $1.47/M to $1.28/M (13% down). (details)
• GLM 5.2 (max) cached input price moved from $0.30/M to $0.26/M (13% down). (details)
• GLM 5.3 Flash cached input price moved from $0.034/M to $0.030/M (13% down). (details)
• Gemini 3.1 Pro cost per task moved from $1.30/M to $1.44/M (11% up). (details)
• Grok 4.7 (xhigh) cost per task moved from $3.74/M to $3.24/M (13% down). (details)
• Muse Spark 1.1 (xhigh) cost per task moved from $1.38/M to $1.69/M (22% up). (details)
• DeepSeek V4.1 Flash (max) cost per task moved from $0.48/M to $0.32/M (33% down). (details)
• Grok 4.7 (low) cost per task moved from $1.25/M to $1.03/M (18% down). (details)
• Qwen3.7 (max) cost per task moved from $1.15/M to $0.93/M (19% down). (details)
• Gemini 3.1 Pro (high) cost per task moved from $1.30/M to $1.44/M (11% up). (details)
• Kimi K2.6 (thinking) cost per task moved from $0.80/M to $0.71/M (12% down). (details)
• Grok 4.5 (high) cost per task moved from $0.88/M to $0.76/M (14% down). (details)
• GLM-5 (thinking) cached input price moved from $0.15/M to $0.20/M (34% up). (details)
• GPT-5.5 Instant cost per task moved from $0.69/M to $0.57/M (18% down). (details)
• DeepSeek V4 Flash 0420 (max) cost per task moved from $0.090/M to $0.098/M (10% up). (details)
• MiMo-V2.5-Pro (thinking) cost per task moved from $0.14/M to $0.15/M (7% up). (details)
• GLM 5.1 (thinking) cost per task moved from $1.02/M to $0.75/M (27% down). (details)
• GLM 5.1 (thinking) cached input price moved from $0.19/M to $0.26/M (34% up). (details)
• …and 106 more.

New benchmark results

• MLCR: 1 new result, including Kimi K3.1 Pro (board)

Score revisions

• GPT-6 Astra on LiveBench 2.0: 86.1% → 85.6%. (details)
• GPT-6 Astra on LiveBench Long Agentic: 82.4% → 77.5%. (details)
• GPT-6 Astra on LiveBench Browser: 84% → 91.3%. (details)
• Kimi K3.1 Pro on LiveBench 2.0: 76.7% → 71.3%. (details)
• Kimi K3.1 Pro on LiveBench Long Agentic: 49.5% → 39.1%. (details)
• Kimi K3.1 Pro on LiveBench Browser: 80% → 75.9%. (details)
• Claude Fable 5.1 on LiveBench 2.0: 88% → 85.7%. (details)
• Claude Fable 5.1 on LiveBench Long Agentic: 80.8% → 77.9%. (details)
• Claude Fable 5.1 on LiveBench Browser: 100% → 89.8%. (details)
• Muse Wonder on LiveBench 2.0: 78.1% → 74.2%. (details)
• Muse Wonder on LiveBench Long Agentic: 53.2% → 52.6%. (details)
• Muse Wonder on LiveBench Browser: 72% → 60%. (details)
• MiMo-V2.6-Pro on LiveBench 2.0: 74.9% → 68.5%. (details)
• MiMo-V2.6-Pro on LiveBench Long Agentic: 47.5% → 35.2%. (details)
• MiMo-V2.6-Pro on LiveBench Browser: 88% → 77.3%. (details)
• Claude Opus 5.5 on LiveBench 2.0: 86.5% → 85.1%. (details)
• Claude Opus 5.5 on LiveBench Long Agentic: 85% → 72.8%. (details)
• Claude Opus 5.5 on LiveBench Browser: 80% → 88.4%. (details)
• Gemini 4 Argon on LiveBench 2.0: 84.1% → 79.3%. (details)
• Gemini 4 Argon on LiveBench Long Agentic: 69.9% → 61.2%. (details)
• …and 72 more.

Speed changes

• GPT-6.1 Sol time to first answer changed from 162.10 s to 312.87 s. (details)
• GPT-6.1 Sol response time changed from 170.82 s to 320.46 s. (details)
• GPT-6 Astra time to first answer changed from 32.11 s to 62.43 s. (details)
• GPT-6 Astra response time changed from 40.94 s to 70.71 s. (details)
• Claude Fable 5.1 time to first answer changed from 60.95 s to 100.72 s. (details)
• Claude Fable 5.1 response time changed from 68.67 s to 108.69 s. (details)
• Claude Opus 5.5 time to first answer changed from 21.55 s to 33.08 s. (details)
• Claude Opus 5.5 response time changed from 27.75 s to 39.33 s. (details)
• Claude Fable 5 time to first answer changed from 59.81 s to 101.84 s. (details)
• Claude Fable 5 response time changed from 66.98 s to 109.25 s. (details)
• Claude Opus 5 time to first answer changed from 24.09 s to 73.02 s. (details)
• Claude Opus 5 response time changed from 32.90 s to 82.58 s. (details)
• Claude Opus 5 time to first token changed from 2.16 s to 3.06 s. (details)
• GPT-5.6 Sol time to first answer changed from 69.47 s to 107.73 s. (details)
• GPT-5.6 Sol response time changed from 75.09 s to 113.85 s. (details)
• …and 146 more.

Today's top five

1. Claude Opus 5.5 (max) — 72.6
2. Gemini 4 Argon (high) — 70.7
3. Muse Wonder Full (max) — 70.6
4. Claude Fable 5.1 (max) — 70.5
5. GPT-6 Astra (max) — 70.4

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.

Head to head

• Claude Opus 5.5 vs Gemini 4 Argon
• Claude Opus 5.5 vs Muse Wonder Full
• Claude Fable 5.1 vs Claude Opus 5.5