BenchLeader

Claude Fable 5.1 up to #5, GLM-4.7-FlashX price cut (4 Nov)

Claude Fable 5.1 moved #6 → #5; GLM-4.7-FlashX got cheaper. 2 top-10 rank changes, 171 price changes, 339 new benchmark results. Sources linked.

BenchLeader refreshed 1691 models from 27 of 27 data feeds (benchmark publishers, aggregators and price lists). 2 top-10 rank changes, 171 price changes, 339 new or updated benchmark results across 27 data feeds.

Movement in the top 10

  • Claude Fable 5.1 (max) moved from #6 to #5 in the BenchLeader Index. (details)
  • GPT-6.1 Sol (max) moved from #5 to #6 in the BenchLeader Index. (details)

Price changes

  • Qwen3.7 (max) cached input price moved from $0.33/M to $0.50/M (52% up). (details)
  • GLM 5.2 (max) cached input price moved from $0.30/M to $0.26/M (13% down). (details)
  • GLM-5 (thinking) cached input price moved from $0.15/M to $0.20/M (34% up). (details)
  • Qwen3.6 Plus cached input price moved from $0.028/M to $0.050/M (81% up). (details)
  • GLM 5.3 (max) cached input price moved from $0.30/M to $0.26/M (13% down). (details)
  • Qwen3.8 Max (max) cached input price moved from $0.22/M to $0.25/M (13% up). (details)
  • Qwen3.7 Plus cached input price moved from $0.055/M to $0.080/M (45% up). (details)
  • Qwen3.5 Plus (thinking) input price moved from $0.12/M to $0.40/M (248% up). (details)
  • Qwen3.5 Plus (thinking) output price moved from $0.69/M to $2.40/M (249% up). (details)
  • GLM 5.3 Flash cached input price moved from $0.034/M to $0.030/M (13% down). (details)
  • GLM 5.1 input price moved from $0.90/M to $1.40/M (56% up). (details)
  • GLM 5.1 output price moved from $3.59/M to $4.40/M (23% up). (details)
  • GLM 5.1 cached input price moved from $0.19/M to $0.26/M (34% up). (details)
  • GLM 5V Turbo input price moved from $0.75/M to $1.20/M (61% up). (details)
  • GLM 5V Turbo output price moved from $3.29/M to $4.00/M (22% up). (details)
  • GLM 5V Turbo cached input price moved from $0.18/M to $0.24/M (34% up). (details)
  • GLM-5 input price moved from $0.60/M to $1.00/M (67% up). (details)
  • GLM-5 output price moved from $2.69/M to $3.20/M (19% up). (details)
  • GLM-5 cached input price moved from $0.15/M to $0.20/M (34% up). (details)
  • GLM 5.3 Flash (max) cached input price moved from $0.034/M to $0.030/M (13% down). (details)
  • GLM 4.7 input price moved from $0.45/M to $0.60/M (34% up). (details)
  • GLM 4.7 output price moved from $2.09/M to $2.20/M (5% up). (details)
  • GLM 4.7 cached input price moved from $0.090/M to $0.11/M (23% up). (details)
  • Qwen3.5 Flash input price moved from $0.029/M to $0.10/M (245% up). (details)
  • Qwen3.5 Flash output price moved from $0.29/M to $0.40/M (39% up). (details)
  • GLM 5.2 input price moved from $1.20/M to $1.40/M (17% up). (details)
  • GLM 5.2 output price moved from $4.18/M to $4.40/M (5% up). (details)
  • GLM 5.2 cached input price moved from $0.30/M to $0.26/M (13% down). (details)
  • Qwen3 32B input price moved from $0.29/M to $0.70/M (144% up). (details)
  • Qwen3 32B output price moved from $1.15/M to $2.80/M (144% up). (details)
  • …and 141 more.

New benchmark results

  • GameBench 2 by reasoning setting: 5 new results, including Claude Fable 5.1, Claude Fable 5.1, Claude Fable 5.1… (board)
  • AA-Omniscience v1.1: 5 new results, including GPT-6 Sol, Grok 4.6, GPT-5.6 Luna… (board)
  • AA-Omniscience v1.1: accuracy: 5 new results, including GPT-6 Sol, Grok 4.6, GPT-5.6 Luna… (board)
  • AA-Omniscience v1.1: non-hallucination: 5 new results, including GPT-6 Sol, Grok 4.6, GPT-5.6 Luna… (board)
  • FrontierCode 2.0 (AA): 5 new results, including GPT-6 Sol, GPT-5.6 Terra, Grok 4.6… (board)
  • CritPt v2 hard subset (AA): 3 new results, including GPT-6 Sol, GPT-5.6 Terra, Grok 4.6 (board)
  • Terminal-Bench-Science 0.1 (AA): 3 new results, including Gemini 3.1 Flash Lite, Claude Haiku 4.5, K2 Horizon MoVA 36B A4B (board)
  • AA Intelligence Index v5.0.1: 2 new results, including GPT-6 Sol, GPT-5.6 Terra (board)
  • GameBench 2 Overall: 2 new results, including Kimi K3.1 Pro, Muse Spark Open (board)
  • HLE-Diamond (AA): 2 new results, including Gemini 3.1 Flash Lite, Gemma 4 26B A4B (board)
  • SREBench: 1 new result, including Kimi K3.1 Pro (board)
  • GPQA Diamond: 1 new result, including Seed 1.6 (board)

Score revisions

  • Claude Opus 5.5 (max) on AA Intelligence Index v5.0: 58.4 → 59.1. (details)
  • Claude Opus 5.5 (max) on CritPt v2 hard subset (AA): 61% → 68%. (details)
  • Gemini 4 Carbon (high) on AA Intelligence Index v5.0: 59.4 → 60. (details)
  • Gemini 4 Carbon (high) on CritPt v2 hard subset (AA): 63% → 69%. (details)
  • GPT-6 Astra on GameBench 2 Overall: 89.7 → 88.7. (details)
  • Claude Fable 5.1 (max) on CritPt v2 hard subset (AA): 51% → 52.5%. (details)
  • GPT-6.1 Sol (max) on CritPt v2 hard subset (AA): 68.5% → 69.5%. (details)
  • GPT-6 Astra (xhigh) on CritPt v2 hard subset (AA): 70.5% → 71%. (details)
  • Claude Fable 5 on GameBench 2 Overall: 67.6 → 65.8. (details)
  • Claude Fable 5.1 on GameBench 2 Overall: 88.6 → 87.2. (details)
  • Claude Opus 5.5 (xhigh) on CritPt v2 hard subset (AA): 63% → 67.5%. (details)
  • Gemini 4 Argon (high) on GameBench 2 by reasoning setting: 90.1 → 89. (details)
  • Claude Opus 5.5 (high) on CritPt v2 hard subset (AA): 58% → 59.5%. (details)
  • GPT-5.6 Sol (max) on CritPt v2 hard subset (AA): 56% → 60%. (details)
  • Claude Sonnet 5.5 (max) on CritPt v2 hard subset (AA): 56% → 59%. (details)
  • GPT-6.1 Terra (max) on CritPt v2 hard subset (AA): 64.5% → 68%. (details)
  • Claude Opus 5 on GameBench 2 Overall: 78.1 → 76.6. (details)
  • Claude Fable 5.1 (xhigh) on CritPt v2 hard subset (AA): 56.5% → 60%. (details)
  • GPT-6 Astra (medium) on CritPt v2 hard subset (AA): 71% → 72%. (details)
  • GPT-6 Astra (medium) on GameBench 2 by reasoning setting: 89.7 → 88.6. (details)
  • …and 278 more.

Speed changes

  • Gemini 4 Carbon time to first answer changed from 29.12 s to 17.75 s. (details)
  • Muse Spark 1.3 time to first token changed from 2.53 s to 3.53 s. (details)
  • Muse Spark 1.3 thinking time changed from 5.98 s to 8.28 s. (details)
  • Grok 4.20 Multi-Agent time to first token changed from 7.65 s to 4.96 s. (details)
  • Qwen3.5 Plus time to first token changed from 1.21 s to 1.67 s. (details)
  • Claude Sonnet 5 time to first answer changed from 28.39 s to 40.76 s. (details)
  • Claude Sonnet 5 response time changed from 35.59 s to 48.13 s. (details)
  • Beam 1.0 501B A23B time to first token changed from 0.40 s to 0.23 s. (details)
  • Mistral Large 4 output speed changed from 170 tok/s to 101 tok/s. (details)
  • Mistral Large 4 time to first answer changed from 13.72 s to 21.81 s. (details)
  • Mistral Large 4 response time changed from 16.66 s to 26.75 s. (details)
  • Mistral Large 4 thinking time changed from 11.76 s to 19.79 s. (details)
  • o4-mini output speed changed from 101 tok/s to 146 tok/s. (details)
  • Qwen3.5-35B-A3B output speed changed from 25 tok/s to 38 tok/s. (details)
  • K2 Horizon 375B A23B output speed changed from 665 tok/s to 123 tok/s. (details)
  • …and 21 more.

Today's top five

  1. Claude Fable 5.2 (max) — 72.7
  2. Gemini 4 Carbon (high) — 68.9
  3. Claude Opus 5.5 (max) — 68.4
  4. GPT-6 Astra (max) — 67.3
  5. 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.

Head to head

Get this in your inbox

No ads, no tracking pixels, unsubscribe in one click.
What to receive