BenchLeader

Daily digest, 13 Sept 2026

93 new models, 7 top-10 rank changes, 22 price changes, 33 new or updated benchmark results across 20 sources.

BenchLeader refreshed 1262 models from 20 of 20 sources. 93 new models, 7 top-10 rank changes, 22 price changes, 33 new or updated benchmark results across 20 sources.

New models

  • AU Anthropic Claude Opus 4.6 (Amazon) is now listed at $33.00/M blended; no independent results yet. (details)
  • AU Anthropic Claude Sonnet 4.6 (Amazon) is now listed at $6.60/M blended; no independent results yet. (details)
  • Claude Fable 5 (EU) (Amazon) is now listed at $22.00/M blended; no independent results yet. (details)
  • Claude Fable 5 (Global) (Amazon) is now listed at $20.00/M blended; no independent results yet. (details)
  • Claude Fable 5 (US) (Amazon) is now listed at $20.00/M blended; no independent results yet. (details)
  • Claude Fable 5.1 (Global) (Amazon) is now listed at $20.00/M blended; no independent results yet. (details)
  • Claude Fable 5.1 (US) (Amazon) is now listed at $22.00/M blended; no independent results yet. (details)
  • Claude Haiku 4.5 (AU) (Amazon) is now listed at $2.00/M blended; no independent results yet. (details)
  • Claude Haiku 4.5 (EU) (Amazon) is now listed at $2.20/M blended; no independent results yet. (details)
  • Claude Haiku 4.5 (Global) (Amazon) is now listed at $2.00/M blended; no independent results yet. (details)
  • Claude Haiku 4.5 (JP) (Amazon) is now listed at $2.00/M blended; no independent results yet. (details)
  • Claude Haiku 4.5 (US) (Amazon) is now listed at $2.00/M blended; no independent results yet. (details)
  • Claude Opus 4.1 (US) (Amazon) is now listed at $30.00/M blended; no independent results yet. (details)
  • Claude Opus 4.5 (EU) (Amazon) is now listed at $11.00/M blended; no independent results yet. (details)
  • Claude Opus 4.5 (Global) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.5 (US) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.6 (EU) (Amazon) is now listed at $11.00/M blended; no independent results yet. (details)
  • Claude Opus 4.6 (Global) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.6 (US) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.7 (AU) (Amazon) is now listed at $11.00/M blended; no independent results yet. (details)
  • Claude Opus 4.7 (EU) (Amazon) is now listed at $11.00/M blended; no independent results yet. (details)
  • Claude Opus 4.7 (Global) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.7 (JP) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.7 (US) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.8 (AU) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.8 (EU) (Amazon) is now listed at $11.00/M blended; no independent results yet. (details)
  • Claude Opus 4.8 (Global) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.8 (JP) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 4.8 (US) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • Claude Opus 5 (AU) (Amazon) is now listed at $10.00/M blended; no independent results yet. (details)
  • …and 63 more.

Movement in the top 10

  • GPT-6 Astra (max) moved from #2 to #1 in the BenchLeader Index. (details)
  • Claude Fable 5.1 (max) moved from #5 to #2 in the BenchLeader Index. (details)
  • Claude Fable 5.1 moved from #1 to #3 in the BenchLeader Index. (details)
  • GPT-6 Astra (high) moved from #6 to #5 in the BenchLeader Index. (details)
  • Claude Fable 5.1 (xhigh) moved from #3 to #6 in the BenchLeader Index. (details)
  • Claude Opus 5 moved from #9 to #8 in the BenchLeader Index. (details)
  • Claude Fable 5.1 (high) moved from #8 to #9 in the BenchLeader Index. (details)

Price changes

  • Gemma 4 26B A4B input price moved from $0.12/M to $0.13/M (8% up). (details)
  • Grok 4 Fast (thinking) input price moved from $0.20/M to $1.25/M (525% up). (details)
  • Grok 4 Fast (thinking) output price moved from $0.50/M to $2.50/M (400% up). (details)
  • Gemma 4 26B A4B (no reasoning) input price moved from $0.12/M to $0.13/M (8% up). (details)
  • Grok 4 Fast (no reasoning) input price moved from $0.20/M to $1.25/M (525% up). (details)
  • Grok 4 Fast (no reasoning) output price moved from $0.50/M to $2.50/M (400% up). (details)
  • Llama 3-70B input price moved from $0.65/M to $0.51/M (22% down). (details)
  • Llama 3-70B output price moved from $2.75/M to $0.74/M (73% down). (details)
  • Llama 3.1 70B input price moved from $0.72/M to $0.56/M (22% down). (details)
  • Llama 3.1 70B output price moved from $0.72/M to $0.56/M (22% down). (details)
  • Nova Micro input price moved from $0.04/M to $0.04/M (12% down). (details)
  • Llama 3-8B input price moved from $0.05/M to $0.04/M (20% down). (details)
  • Llama 3-8B output price moved from $0.14/M to $0.04/M (71% down). (details)
  • Mixtral 8x7B input price moved from $0.45/M to $0.00/M (100% down). (details)
  • Mixtral 8x7B output price moved from $0.70/M to $0.00/M (100% down). (details)
  • Gemma 4 26B A4B (minimal) input price moved from $0.12/M to $0.13/M (8% up). (details)
  • Muse Glimmer 30B input price moved from $0.30/M to $0.35/M (17% up). (details)
  • Muse Glimmer 30B output price moved from $1.10/M to $1.50/M (36% up). (details)
  • Nemotron 3 Nano Omni 30B A3b (thinking) input price moved from $0.09/M to $0.20/M (122% up). (details)
  • Nemotron 3 Nano Omni 30B A3b (thinking) output price moved from $0.36/M to $0.80/M (122% up). (details)
  • Reka Flash 3 input price moved from $0.20/M to $0.10/M (50% down). (details)
  • Reka Flash 3 output price moved from $0.80/M to $0.20/M (75% down). (details)

New benchmark results

  • AA Intelligence Index: 1 new result, including Granite 3.3 8B (board)
  • GPQA Diamond (AA): 1 new result, including Granite 3.3 8B (board)
  • Humanity's Last Exam (AA): 1 new result, including Granite 3.3 8B (board)
  • IFBench: 1 new result, including Granite 3.3 8B (board)
  • τ²-Bench Telecom (AA): 1 new result, including Granite 3.3 8B (board)
  • Terminal-Bench Hard: 1 new result, including Granite 3.3 8B (board)
  • AA-Omniscience: 1 new result, including Granite 3.3 8B (board)

Score revisions

  • GPT-5.5 Instant on GPQA Diamond (AA): 84.7% → 82.3%. (details)
  • GPT-5.5 Instant on Humanity's Last Exam (AA): 21.6% → 19.9%. (details)
  • GPT-5.5 Instant on AA-Omniscience: 11.2 → 3.7. (details)
  • GLM 5.2 on AA Intelligence Index: 38.6 → 34. (details)
  • Grok 4.20 on IFBench: 82.9% → 81.2%. (details)
  • Grok 4.20 on τ²-Bench Telecom (AA): 96.5% → 93%. (details)
  • Grok 4.20 on Terminal-Bench Hard: 40.9% → 37.9%. (details)
  • Gemini 2.5 Flash on AA Intelligence Index: 11.7 → 9.9. (details)
  • Gemini 2.5 Flash on GPQA Diamond (AA): 69.8% → 68.3%. (details)
  • Gemini 2.5 Flash on Humanity's Last Exam (AA): 12.1% → 4.7%. (details)
  • Qwen3 (max) on AA Intelligence Index: 21.3 → 16.3. (details)
  • Qwen3 (max) on GPQA Diamond (AA): 86.1% → 77.6%. (details)
  • Qwen3 (max) on Humanity's Last Exam (AA): 28% → 12.7%. (details)
  • Qwen3 (max) on IFBench: 70.8% → 53.8%. (details)
  • Qwen3 (max) on AA-LCR: 74.3% → 62%. (details)
  • Qwen3 (max) on Terminal-Bench Hard: 24.2% → 17.4%. (details)
  • Qwen3 (max) on AA-Omniscience: -36.6 → -39.1. (details)
  • Grok 4.20 (no reasoning) on GPQA Diamond (AA): 78.5% → 77.6%. (details)
  • Grok 4.20 (no reasoning) on AA-LCR: 25.3% → 23.3%. (details)
  • Grok 4.20 (no reasoning) on τ²-Bench Telecom (AA): 69.6% → 59.9%. (details)
  • …and 6 more.

Speed changes

  • Nemotron 3 Ultra 550B A55B time to first token changed from 1.79 s to 1.09 s. (details)
  • GPT-5.6 Luna time to first answer changed from 1.50 s to 2.10 s. (details)
  • Nemotron 3 Super time to first token changed from 1.05 s to 1.47 s. (details)
  • Claude Sonnet 4.5 time to first answer changed from 14.89 s to 9.44 s. (details)
  • Qwen3.5-4B output speed changed from 18 tok/s to 26 tok/s. (details)
  • Llama Nemotron Super 49B v1.5 time to first answer changed from 6.42 s to 3.50 s. (details)
  • Llama Nemotron Super 49B v1.5 time to first token changed from 6.42 s to 3.50 s. (details)
  • Nemotron Nano 12B v2 VL output speed changed from 76 tok/s to 121 tok/s. (details)
  • Claude Sonnet 5 time to first answer changed from 1.25 s to 0.75 s. (details)
  • Claude Sonnet 5 time to first token changed from 1.25 s to 0.75 s. (details)
  • Mistral Small 3 output speed changed from 9 tok/s to 26 tok/s. (details)
  • Lyria 3 Pro output speed changed from 2 tok/s to 3 tok/s. (details)

Today's top five

  1. GPT-6 Astra — 70.5
  2. Claude Fable 5.1 — 70.1
  3. Claude Fable 5.1 — 69.1
  4. Claude Fable 5 — 68.8
  5. GPT-6 Astra — 68.7

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.

Get this in your inbox

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