Daily digest, 10 Oct 2026

BenchLeader refreshed 1560 models from 23 of 23 sources. 14 price changes, 50 new or updated benchmark results across 23 sources.

Price changes

• Qwen3.8 27B input price moved from $0.45/M to $0.42/M (6% down). (details)
• DeepSeek V4 Flash 0731 (high) cached input price moved from $0.03/M to $0.03/M (7% down). (details)
• DeepSeek V4 Flash 0731 (max) cached input price moved from $0.03/M to $0.03/M (7% down). (details)
• DeepSeek V4 Flash 0731 cached input price moved from $0.03/M to $0.03/M (7% down). (details)
• LongCat 2.0 input price moved from $0.30/M to $0.75/M (150% up). (details)
• LongCat 2.0 output price moved from $1.20/M to $2.95/M (146% up). (details)
• LongCat 2.0 cost per task moved from 0.1 to 0.1 (148% up). (details)
• Nemotron 3 Nano 30B A3B input price moved from $0.06/M to $0.05/M (17% down). (details)
• Nemotron 3 Nano 30B A3B output price moved from $0.24/M to $0.20/M (17% down). (details)
• DeepSeek V4 Flash 0731 (no reasoning) cached input price moved from $0.03/M to $0.03/M (7% down). (details)
• DeepSeek V4 Flash 0731 (low) cached input price moved from $0.03/M to $0.03/M (7% down). (details)
• BGE M3 input price moved from $0.00/M to $0.01/M (1000000000% up). (details)
• Ling 3.0 Flash Sante input price moved from $0.04/M to $0.07/M (79% up). (details)
• Ling 3.0 Flash Sante output price moved from $0.12/M to $0.22/M (79% up). (details)

New benchmark results

• PRBench Finance: 35 new results, including Claude Fable 5, GPT-5.6 Sol, GPT-6 Astra… (board)
• OTIS Mock AIME: 3 new results, including DeepSeek V4.1 Flash, Claude Haiku 5.5, Claude Haiku 5.5 (board)
• GPQA Diamond: 2 new results, including DeepSeek V4.1 Flash, Claude Haiku 5.5 (board)
• FrontierMath Tiers 1–3: 2 new results, including DeepSeek V4.1 Flash, Claude Haiku 5.5 (board)
• FrontierMath Tier 4: 2 new results, including DeepSeek V4.1 Flash, Claude Haiku 5.5 (board)
• Mystery Game Puzzles: 2 new results, including DeepSeek V4.1 Flash, Claude Haiku 5.5 (board)
• SimpleQA Verified: 2 new results, including Claude Haiku 5.5, Mistral Large 4 (board)

Score revisions

• GPT-6.1 Sol (max) on AA-Briefcase v1.1: 1564 → 1557. (details)
• GPT-6.1 Sol (high) on AA-Briefcase v1.1: 1471 → 1465. (details)

Speed changes

• GPT-5.6 Sol time to first token changed from 5.76 s to 3.59 s. (details)
• GPT-5.6 Sol time to first answer changed from 9.70 s to 20.45 s. (details)
• GPT-5.6 Sol response time changed from 17.11 s to 27.03 s. (details)
• GPT-5.5 time to first token changed from 2.23 s to 3.23 s. (details)
• Qwen3.8 2.4T A95B time to first token changed from 0.34 s to 1.48 s. (details)
• GPT-5.5 time to first answer changed from 5.30 s to 7.61 s. (details)
• GLM 5.2 time to first token changed from 0.58 s to 0.89 s. (details)
• GPT-5.4 Pro output speed changed from 40 tok/s to 6 tok/s. (details)
• GPT-5.6 Terra time to first answer changed from 2.64 s to 4.81 s. (details)
• Grok 4.20 time to first token changed from 1.15 s to 0.73 s. (details)
• Gemma 4 26B A4B time to first token changed from 0.87 s to 0.56 s. (details)
• GPT-5 nano time to first token changed from 3.06 s to 1.85 s. (details)
• Muse Glimmer time to first token changed from 1.92 s to 1.12 s. (details)
• Qwen3.5-27B time to first token changed from 1.88 s to 0.91 s. (details)
• Qwen3.5-122B-A10B output speed changed from 45 tok/s to 67 tok/s. (details)
• …and 44 more.

Today's top five

1. Claude Opus 5.5 — 72.0
2. Claude Fable 5.1 — 70.8
3. GPT-6 Astra — 70.2
4. Claude Opus 5.5 — 70.2
5. Gemini 4 Argon — 70.0

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.