How much of a workforce needs a top-tier model DERIVED 14.8% of staff need tier 70 or better, and 0.7% need the top tier. Drawn from 297 roles carrying 17,116 people.
Crossing that line costs 6.7x more per token: the cheapest model clearing tier 30 is Qwen3.5 4B (Reasoning) at $0.03/M, the cheapest clearing tier 70 is GPT-5.6 Luna (max) at $0.20/M.
Industry All industries * Accounting & Audit Aerospace & Defence Agriculture & Food Production Airlines & Aviation Automotive Banking Biotechnology Cloud & Digital Infrastructure Construction & Engineering Consumer Goods Education Gaming & Interactive Entertainment Healthcare Providers Higher Education & Research IT Services & Consulting Insurance Investment Banking & Capital Markets Legal Services Management Consulting Manufacturing Media & Entertainment Medical Devices Mining & Metals Oil & Gas Payments & FinTech Pharmaceuticals Power & Utilities Public Sector & Government Real Estate & Property Services Renewable Energy Retail & E-commerce Software & SaaS Telecommunications Transport & Logistics Travel, Hospitality & Leisure Wealth & Asset Management View Smoothed curve Measured tiers
Capability General intelligence Specialist knowledge· Instruction following· Tool use·
Tier 10: 5.8% of staff (1,000 people) Tier 30: 39.8% of staff (6,804 people) Tier 50: 39.6% of staff (6,776 people) Tier 70: 14.1% of staff (2,415 people) Tier 90: 0.7% of staff (121 people) 14.8% of staff need this or better density 0 tier 10 20 tier 30 32 tier 50 45 tier 70 56 tier 90 $/M in Intelligence Index v4.1 (Artificial Analysis) This is the shape of a reference 17,116-person enterprise, not your org chart. The headcounts are per-role reference figures published with the role library. Upload your own headcount to see yours.
The roles that do need it, largest first Role Industry Function People Tier Actuarial Analyst Insurance Actuarial 60 70 Bioinformatics Scientist Biotechnology Bioinformatics & Computational Biology 60 70 Capital Markets Associate Investment Banking & Capital Markets Capital Markets Origination 60 70 Corporate Lawyer Legal Services Corporate & Transactional Law 60 70 Curriculum Designer Education Curriculum & Programme Design 60 70 Data Scientist * Data, Analytics & AI 60 70 Editor Media & Entertainment Content Development & Editorial 60 70 Emergency Planning Officer Public Sector & Government Emergency, Justice & National Security Operations 60 70 Equity Research Analyst Investment Banking & Capital Markets Research 60 70 Forensic Accountant Accounting & Audit Forensic & Investigations 60 70 Forward Deployed Engineer Software & SaaS Forward Deployed Engineering 60 70 Forward Deployed Engineer IT Services & Consulting Forward Deployed Engineering 60 70
How this is derived
What capability costs AIE dataset General intelligence Multi-step reasoning Agentic Accuracy Latency and speed Coding· no data
General intelligence: 330 of 330 scored. Cheaper is to the left, on a log scale.
$0.02 $0.10 $1 $10 $50 Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)
General intelligence: 60.1
$5/M input
Dominated: something cheaper scores at least as well Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
General intelligence: 59.9
$10/M input
Dominated: something cheaper scores at least as well Claude Opus 5 (Adaptive Reasoning, High Effort)
General intelligence: 58.9
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (max)
General intelligence: 58.9
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (xhigh)
General intelligence: 57.7
$5/M input
Dominated: something cheaper scores at least as well Claude Opus 5 (Adaptive Reasoning, Medium Effort)
General intelligence: 56.3
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (high)
General intelligence: 55.9
$5/M input
Dominated: something cheaper scores at least as well Claude Opus 4.8 (Adaptive Reasoning, Max Effort)
General intelligence: 55.7
$5/M input
Dominated: something cheaper scores at least as well GPT-5.5 (xhigh)
General intelligence: 54.8
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (medium)
General intelligence: 53.6
$5/M input
Dominated: something cheaper scores at least as well Claude Opus 4.7 (Adaptive Reasoning, Max Effort)
General intelligence: 53.5
$5/M input
Dominated: something cheaper scores at least as well Claude Sonnet 5 (Adaptive Reasoning, Max Effort)
General intelligence: 53.4
$2/M input
Dominated: something cheaper scores at least as well GPT-5.5 (high)
General intelligence: 53.1
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Terra (xhigh)
General intelligence: 51.6
$2/M input
Dominated: something cheaper scores at least as well GPT-5.4 (xhigh)
General intelligence: 51.4
$3/M input
Dominated: something cheaper scores at least as well GLM-5.2 (max)
General intelligence: 51.1
$1/M input
Dominated: something cheaper scores at least as well Muse Spark 1.1 (xhigh)
General intelligence: 50.6
$1/M input
Dominated: something cheaper scores at least as well Claude Opus 5 (Adaptive Reasoning, Low Effort)
General intelligence: 50.6
$5/M input
Dominated: something cheaper scores at least as well GPT-5.5 (medium)
General intelligence: 50.4
$5/M input
Dominated: something cheaper scores at least as well Gemini 3.5 Flash (high)
General intelligence: 50.2
$2/M input
Dominated: something cheaper scores at least as well Gemini 3.6 Flash (high)
General intelligence: 50.1
$2/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (low)
General intelligence: 49.4
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Luna (xhigh)
General intelligence: 49.1
$0.20/M input
Dominated: something cheaper scores at least as well GPT-5.6 Terra (high)
General intelligence: 49
$2/M input
Dominated: something cheaper scores at least as well Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)
General intelligence: 47.2
$3/M input
Dominated: something cheaper scores at least as well Gemini 3.1 Pro Preview
General intelligence: 46.5
$2/M input
Dominated: something cheaper scores at least as well GPT-5.6 Luna (high)
General intelligence: 46.1
$1/M input
Dominated: something cheaper scores at least as well Qwen3.7 Max
General intelligence: 46
$3/M input
Dominated: something cheaper scores at least as well GPT-5.6 Terra (medium)
General intelligence: 45.6
$2/M input
Dominated: something cheaper scores at least as well Gemini 3.5 Flash (medium)
General intelligence: 45.4
$2/M input
Dominated: something cheaper scores at least as well MiniMax-M3
General intelligence: 44.4
$0.30/M input
Dominated: something cheaper scores at least as well DeepSeek V4 Pro (Reasoning, Max Effort)
General intelligence: 44.3
$0.43/M input
Dominated: something cheaper scores at least as well GPT-5.3 Codex (xhigh)
General intelligence: 44.3
$2/M input
Dominated: something cheaper scores at least as well Kimi K2.6
General intelligence: 44.2
$0.95/M input
Dominated: something cheaper scores at least as well Claude Opus 4.6 (Adaptive Reasoning, Max Effort)
General intelligence: 43.7
$5/M input
Dominated: something cheaper scores at least as well GPT-5.5 (low)
General intelligence: 43.5
$5/M input
Dominated: something cheaper scores at least as well DeepSeek V4 Pro (Reasoning, High Effort)
General intelligence: 43.1
$0.43/M input
Dominated: something cheaper scores at least as well Claude Opus 4.7 (Non-reasoning, High Effort)
General intelligence: 42.7
$5/M input
Dominated: something cheaper scores at least as well GPT-5.2 (xhigh)
General intelligence: 42.2
$2/M input
Dominated: something cheaper scores at least as well Kimi K2.7 Code
General intelligence: 41.9
$0.95/M input
Dominated: something cheaper scores at least as well Claude Sonnet 5 (Non-reasoning, High Effort)
General intelligence: 41.7
$2/M input
Dominated: something cheaper scores at least as well GPT-5.6 Sol (Non-reasoning)
General intelligence: 41.2
$5/M input
Dominated: something cheaper scores at least as well Claude Opus 4.5 (Reasoning)
General intelligence: 40.8
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Terra (low)
General intelligence: 40.5
$3/M input
Dominated: something cheaper scores at least as well GLM-5.1 (Reasoning)
General intelligence: 40.2
$1/M input
Dominated: something cheaper scores at least as well GPT-5.2 Codex (xhigh)
General intelligence: 40.1
$2/M input
Dominated: something cheaper scores at least as well GPT-5.4 mini (xhigh)
General intelligence: 40
$0.75/M input
Dominated: something cheaper scores at least as well Qwen3.6 Max Preview
General intelligence: 40
$1/M input
Dominated: something cheaper scores at least as well Grok Build 0.1 0616
General intelligence: 39.8
$1/M input
Dominated: something cheaper scores at least as well Qwen3.6 Plus
General intelligence: 39.6
$0.50/M input
Dominated: something cheaper scores at least as well Gemini 3 Pro Preview (high)
General intelligence: 39.6
$2/M input
Dominated: something cheaper scores at least as well GLM-5 (Reasoning)
General intelligence: 39.5
$1/M input
Dominated: something cheaper scores at least as well GPT-5.4 (low)
General intelligence: 39.1
$3/M input
Dominated: something cheaper scores at least as well Qwen3.7 Plus
General intelligence: 39
$0.40/M input
Dominated: something cheaper scores at least as well GPT-5.4 nano (xhigh)
General intelligence: 38.2
$0.20/M input
Dominated: something cheaper scores at least as well GPT-5.6 Luna (medium)
General intelligence: 38.1
$0.20/M input
Dominated: something cheaper scores at least as well MiniMax-M2.7
General intelligence: 38.1
$0.30/M input
Dominated: something cheaper scores at least as well GPT-5.2 (medium)
General intelligence: 38
$2/M input
Dominated: something cheaper scores at least as well Gemini 3 Flash Preview (Reasoning)
General intelligence: 37.8
$0.50/M input
Dominated: something cheaper scores at least as well Nemotron 3 Ultra 550B A55B (Reasoning)
General intelligence: 37.8
$0.68/M input
Dominated: something cheaper scores at least as well Claude Opus 4.6 (Non-reasoning, High Effort)
General intelligence: 37.8
$5/M input
Dominated: something cheaper scores at least as well Grok 4.3 (high)
General intelligence: 37.6
$1/M input
Dominated: something cheaper scores at least as well DeepSeek V4 Flash (Reasoning, High Effort)
General intelligence: 37.5
$0.14/M input
Dominated: something cheaper scores at least as well Qwen3.6 27B (Reasoning)
General intelligence: 37.1
$0.60/M input
Dominated: something cheaper scores at least as well Grok 4.20 0309 v2 (Reasoning)
General intelligence: 37
$2/M input
Dominated: something cheaper scores at least as well GPT-5.1 (high)
General intelligence: 36.9
$1/M input
Dominated: something cheaper scores at least as well Gemini 3.5 Flash-Lite
General intelligence: 36.5
$0.30/M input
Dominated: something cheaper scores at least as well Grok 4.20 0309 (Reasoning)
General intelligence: 36.5
$2/M input
Dominated: something cheaper scores at least as well Claude 4.5 Sonnet (Reasoning)
General intelligence: 36.4
$3/M input
Dominated: something cheaper scores at least as well GPT-5 Codex (high)
General intelligence: 36.1
$1/M input
Dominated: something cheaper scores at least as well Grok 4.3 (medium)
General intelligence: 36
$1/M input
Dominated: something cheaper scores at least as well Claude Sonnet 4.6 (Non-reasoning, High Effort)
General intelligence: 35.9
$3/M input
Dominated: something cheaper scores at least as well Kimi K2.5 (Reasoning)
General intelligence: 35.4
$0.60/M input
Dominated: something cheaper scores at least as well Grok 4.3 (low)
General intelligence: 35.4
$1/M input
Dominated: something cheaper scores at least as well GLM-5.1 (Non-reasoning)
General intelligence: 35.4
$1/M input
Dominated: something cheaper scores at least as well GPT-5.5 (Non-reasoning)
General intelligence: 35.4
$5/M input
Dominated: something cheaper scores at least as well Gemini 3.5 Flash (minimal)
General intelligence: 34.9
$2/M input
Dominated: something cheaper scores at least as well GPT-5.1 Codex (high)
General intelligence: 34.7
$1/M input
Dominated: something cheaper scores at least as well GPT-5 (high)
General intelligence: 34.7
$1/M input
Dominated: something cheaper scores at least as well Claude Opus 4.5 (Non-reasoning)
General intelligence: 34.7
$5/M input
Dominated: something cheaper scores at least as well Kimi K2.6 (Non-reasoning)
General intelligence: 34.6
$0.95/M input
Dominated: something cheaper scores at least as well Claude Sonnet 4.6 (Non-reasoning, Low Effort)
General intelligence: 34.3
$3/M input
Dominated: something cheaper scores at least as well GLM-5.2 (Non-reasoning)
General intelligence: 34.1
$1/M input
Dominated: something cheaper scores at least as well GPT-5.6 Terra (Non-reasoning)
General intelligence: 34
$2/M input
Dominated: something cheaper scores at least as well Qwen3.5 27B (Reasoning)
General intelligence: 33.8
$0.30/M input
Dominated: something cheaper scores at least as well MiniMax-M2.5
General intelligence: 33.7
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3.5 397B A17B (Reasoning)
General intelligence: 33.7
$0.60/M input
Dominated: something cheaper scores at least as well GLM-4.7 (Reasoning)
General intelligence: 33.7
$0.60/M input
Dominated: something cheaper scores at least as well GPT-5 (medium)
General intelligence: 33.7
$1/M input
Dominated: something cheaper scores at least as well Claude 4.1 Opus (Reasoning)
General intelligence: 33.7
$15/M input
Dominated: something cheaper scores at least as well GPT-5.5 Instant (May 2026)
General intelligence: 33.5
$5/M input
Dominated: something cheaper scores at least as well GPT-5.6 Luna (low)
General intelligence: 33.3
$0.20/M input
Dominated: something cheaper scores at least as well Grok 4
General intelligence: 33.3
$3/M input
Dominated: something cheaper scores at least as well Gemini 3 Pro Preview (low)
General intelligence: 33.1
$2/M input
Dominated: something cheaper scores at least as well Kimi K2 Thinking
General intelligence: 32.7
$0.60/M input
Dominated: something cheaper scores at least as well o3-pro
General intelligence: 32.5
$20/M input
Dominated: something cheaper scores at least as well GLM-5 (Non-reasoning)
General intelligence: 32.4
$1/M input
Dominated: something cheaper scores at least as well Qwen3.5 122B A10B (Reasoning)
General intelligence: 32.3
$0.40/M input
Dominated: something cheaper scores at least as well DeepSeek V3.2 (Reasoning)
General intelligence: 32
$0.28/M input
Dominated: something cheaper scores at least as well Qwen3.5 397B A17B (Non-reasoning)
General intelligence: 32
$0.60/M input
Dominated: something cheaper scores at least as well Qwen3.6 35B A3B (Reasoning)
General intelligence: 31.6
$0.25/M input
Dominated: something cheaper scores at least as well MiniMax-M2.1
General intelligence: 31.4
$0.30/M input
Dominated: something cheaper scores at least as well DeepSeek V4 Pro (Non-reasoning)
General intelligence: 31.2
$0.43/M input
Dominated: something cheaper scores at least as well GPT-5 (low)
General intelligence: 31.2
$1/M input
Dominated: something cheaper scores at least as well Claude 4 Opus (Reasoning)
General intelligence: 31
$15/M input
Dominated: something cheaper scores at least as well GPT-5 mini (medium)
General intelligence: 30.9
$0.25/M input
Dominated: something cheaper scores at least as well GPT-5.1 Codex mini (high)
General intelligence: 30.6
$0.25/M input
Dominated: something cheaper scores at least as well Qwen3.5 Omni Plus
General intelligence: 30.6
$0.40/M input
Dominated: something cheaper scores at least as well Qwen3.6 27B (Non-reasoning)
General intelligence: 30.5
$0.60/M input
Dominated: something cheaper scores at least as well DeepSeek V3.1 Terminus (Reasoning)
General intelligence: 30.4
$2/M input
Dominated: something cheaper scores at least as well o3
General intelligence: 30.4
$2/M input
Dominated: something cheaper scores at least as well GPT-5.4 nano (medium)
General intelligence: 30.2
$0.20/M input
Dominated: something cheaper scores at least as well Mistral Medium 3.5
General intelligence: 29.9
$2/M input
Dominated: something cheaper scores at least as well GPT-5.4 mini (medium)
General intelligence: 29.8
$0.75/M input
Dominated: something cheaper scores at least as well Claude 4.5 Haiku (Reasoning)
General intelligence: 29.6
$1/M input
Dominated: something cheaper scores at least as well Kimi K2.5 (Non-reasoning)
General intelligence: 29.4
$0.60/M input
Dominated: something cheaper scores at least as well Qwen3.5 35B A3B (Reasoning)
General intelligence: 29.3
$0.25/M input
Dominated: something cheaper scores at least as well Qwen3.5 27B (Non-reasoning)
General intelligence: 29.3
$0.30/M input
Dominated: something cheaper scores at least as well Claude 4.5 Sonnet (Non-reasoning)
General intelligence: 29.3
$3/M input
Dominated: something cheaper scores at least as well Claude 4 Sonnet (Reasoning)
General intelligence: 28.9
$3/M input
Dominated: something cheaper scores at least as well GPT-5.5 Instant (June 2026)
General intelligence: 28.9
$5/M input
Dominated: something cheaper scores at least as well DeepSeek V4 Flash (Non-reasoning)
General intelligence: 28.7
$0.14/M input
Dominated: something cheaper scores at least as well GLM-4.6 (Reasoning)
General intelligence: 28.7
$0.55/M input
Dominated: something cheaper scores at least as well MiniMax-M2
General intelligence: 28.3
$0.30/M input
Dominated: something cheaper scores at least as well Claude 4.1 Opus (Non-reasoning)
General intelligence: 28.2
$15/M input
Dominated: something cheaper scores at least as well GPT-5.4 (Non-reasoning)
General intelligence: 27.7
$3/M input
Dominated: something cheaper scores at least as well Qwen3.5 122B A10B (Non-reasoning)
General intelligence: 27.6
$0.40/M input
Dominated: something cheaper scores at least as well Grok 4 Fast (Reasoning)
General intelligence: 27.4
$0.20/M input
Dominated: something cheaper scores at least as well Gemini 3 Flash Preview (Non-reasoning)
General intelligence: 27.4
$0.50/M input
Dominated: something cheaper scores at least as well GLM-4.7 (Non-reasoning)
General intelligence: 26.6
$0.60/M input
Dominated: something cheaper scores at least as well GPT-5.6 Luna (Non-reasoning)
General intelligence: 26.6
$1/M input
Dominated: something cheaper scores at least as well GPT-5.2 (Non-reasoning)
General intelligence: 26
$2/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Pro
General intelligence: 25.8
$1/M input
Dominated: something cheaper scores at least as well o4-mini (high)
General intelligence: 25.6
$1/M input
Dominated: something cheaper scores at least as well Claude 4 Sonnet (Non-reasoning)
General intelligence: 25.5
$3/M input
Dominated: something cheaper scores at least as well Claude 4 Opus (Non-reasoning)
General intelligence: 25.5
$15/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron 3 Super 120B A12B (Reasoning)
General intelligence: 25.4
$0.20/M input
Dominated: something cheaper scores at least as well DeepSeek V3.2 Exp (Reasoning)
General intelligence: 25.4
$0.28/M input
Dominated: something cheaper scores at least as well GPT-5 mini (high)
General intelligence: 25.3
$0.25/M input
Dominated: something cheaper scores at least as well Gemini 3.1 Flash-Lite
General intelligence: 25
$0.25/M input
Dominated: something cheaper scores at least as well Qwen3 Max Thinking (Preview)
General intelligence: 25
$1/M input
Dominated: something cheaper scores at least as well Grok 4.3 (Non-reasoning)
General intelligence: 24.8
$1/M input
Dominated: something cheaper scores at least as well DeepSeek V3.2 (Non-reasoning)
General intelligence: 24.7
$0.28/M input
Dominated: something cheaper scores at least as well Qwen3.6 35B A3B (Non-reasoning)
General intelligence: 24.2
$0.38/M input
Dominated: something cheaper scores at least as well Qwen3.5 35B A3B (Non-reasoning)
General intelligence: 24
$0.25/M input
Dominated: something cheaper scores at least as well Qwen3 Max
General intelligence: 24
$1/M input
Dominated: something cheaper scores at least as well gpt-oss-120b (high)
General intelligence: 23.8
$0.15/M input
Dominated: something cheaper scores at least as well Claude 4.5 Haiku (Non-reasoning)
General intelligence: 23.7
$1/M input
Dominated: something cheaper scores at least as well Kimi K2 0905
General intelligence: 23.5
$0.60/M input
Dominated: something cheaper scores at least as well Claude 3.7 Sonnet (Non-reasoning)
General intelligence: 23.5
$3/M input
Dominated: something cheaper scores at least as well o1
General intelligence: 23.4
$15/M input
Dominated: something cheaper scores at least as well GLM-4.6 (Non-reasoning)
General intelligence: 23
$0.57/M input
Dominated: something cheaper scores at least as well Grok 3 mini Reasoning (high)
General intelligence: 22.5
$0.30/M input
Dominated: something cheaper scores at least as well Grok 4.20 0309 (Non-reasoning)
General intelligence: 22.5
$2/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Pro Preview (May' 25)
General intelligence: 22.3
$1/M input
Dominated: something cheaper scores at least as well Gemma 4 12B (Reasoning)
General intelligence: 21.8
$0.10/M input
Dominated: something cheaper scores at least as well Gemma 4 31B (Non-reasoning)
General intelligence: 21.8
$0.14/M input
Dominated: something cheaper scores at least as well Nova 2.0 Pro Preview (medium)
General intelligence: 21.8
$1/M input
Dominated: something cheaper scores at least as well Grok 4.20 0309 v2 (Non-reasoning)
General intelligence: 21.8
$2/M input
Dominated: something cheaper scores at least as well Qwen3.5 9B (Reasoning)
General intelligence: 21.4
$0.14/M input
Dominated: something cheaper scores at least as well DeepSeek V3.1 Terminus (Non-reasoning)
General intelligence: 21.4
$0.27/M input
Dominated: something cheaper scores at least as well DeepSeek V3.2 Exp (Non-reasoning)
General intelligence: 21.3
$0.28/M input
Dominated: something cheaper scores at least as well Qwen3 Coder Next
General intelligence: 21.1
$0.35/M input
Dominated: something cheaper scores at least as well DeepSeek V3.1 (Non-reasoning)
General intelligence: 21
$0.56/M input
Dominated: something cheaper scores at least as well Nova 2.0 Omni (medium)
General intelligence: 20.9
$0.30/M input
Dominated: something cheaper scores at least as well DeepSeek V3.1 (Reasoning)
General intelligence: 20.7
$0.59/M input
Dominated: something cheaper scores at least as well Qwen3 VL 235B A22B (Reasoning)
General intelligence: 20.6
$0.70/M input
Dominated: something cheaper scores at least as well GPT-5.1 (Non-reasoning)
General intelligence: 20.4
$1/M input
Dominated: something cheaper scores at least as well Gemma 4 26B A4B (Non-reasoning)
General intelligence: 20.1
$0.13/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash (Reasoning)
General intelligence: 20.1
$0.30/M input
Dominated: something cheaper scores at least as well DeepSeek R1 0528 (May '25)
General intelligence: 20.1
$1/M input
Dominated: something cheaper scores at least as well GPT-5 nano (high)
General intelligence: 19.9
$0.05/M input
Dominated: something cheaper scores at least as well Mistral Small 4 (Reasoning)
General intelligence: 19.6
$0.15/M input
Dominated: something cheaper scores at least as well Qwen3 235B A22B 2507 (Reasoning)
General intelligence: 19.6
$0.70/M input
Dominated: something cheaper scores at least as well Nova 2.0 Pro Preview (low)
General intelligence: 19.6
$1/M input
Dominated: something cheaper scores at least as well Kimi K2
General intelligence: 19.4
$0.57/M input
Dominated: something cheaper scores at least as well GPT-4.1
General intelligence: 19.4
$2/M input
Dominated: something cheaper scores at least as well Qwen3 Max (Preview)
General intelligence: 19.2
$1/M input
Dominated: something cheaper scores at least as well GPT-5 nano (medium)
General intelligence: 19
$0.05/M input
Dominated: something cheaper scores at least as well Qwen3.5 Omni Flash
General intelligence: 19
$0.10/M input
Dominated: something cheaper scores at least as well Nova 2.0 Lite (medium)
General intelligence: 19
$0.30/M input
Dominated: something cheaper scores at least as well o3-mini
General intelligence: 19
$1/M input
Dominated: something cheaper scores at least as well o1-pro
General intelligence: 18.9
$150/M input
Dominated: something cheaper scores at least as well DeepSeek R1 (Jan '25)
General intelligence: 18.5
$2/M input
Dominated: something cheaper scores at least as well Grok 3
General intelligence: 18.4
$4/M input
Dominated: something cheaper scores at least as well Nova 2.0 Lite (high)
General intelligence: 18.2
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3 235B A22B 2507 Instruct
General intelligence: 18.2
$0.70/M input
Dominated: something cheaper scores at least as well Qwen3 Coder 480B A35B Instruct
General intelligence: 18
$2/M input
Dominated: something cheaper scores at least as well Qwen3 VL 32B (Reasoning)
General intelligence: 17.9
$0.70/M input
Dominated: something cheaper scores at least as well Magistral Medium 1.2
General intelligence: 17.9
$2/M input
Dominated: something cheaper scores at least as well Nova 2.0 Lite (low)
General intelligence: 17.8
$0.30/M input
Dominated: something cheaper scores at least as well MiniMax M1 80k
General intelligence: 17.7
$0.55/M input
Dominated: something cheaper scores at least as well GPT-5.4 nano (Non-Reasoning)
General intelligence: 17.6
$0.20/M input
Dominated: something cheaper scores at least as well GPT-5 (minimal)
General intelligence: 17.2
$1/M input
Dominated: something cheaper scores at least as well o1-preview
General intelligence: 17
$17/M input
Dominated: something cheaper scores at least as well GLM-4.6V (Reasoning)
General intelligence: 16.8
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3 Next 80B A3B (Reasoning)
General intelligence: 16.7
$0.50/M input
Dominated: something cheaper scores at least as well Nova 2.0 Omni (low)
General intelligence: 16.6
$0.30/M input
Dominated: something cheaper scores at least as well GPT-5.4 mini (Non-Reasoning)
General intelligence: 16.6
$0.75/M input
Dominated: something cheaper scores at least as well GLM-4.5-Air
General intelligence: 16.5
$0.17/M input
Dominated: something cheaper scores at least as well Grok 4 Fast (Non-reasoning)
General intelligence: 16.5
$0.20/M input
Dominated: something cheaper scores at least as well Qwen3.5 4B (Non-reasoning)
General intelligence: 16
$0.03/M input
Dominated: something cheaper scores at least as well Mistral Large 3
General intelligence: 15.9
$0.50/M input
Dominated: something cheaper scores at least as well o3-mini (high)
General intelligence: 15.6
$1/M input
Dominated: something cheaper scores at least as well GLM-4.7-Flash (Non-reasoning)
General intelligence: 15.5
$0.07/M input
Dominated: something cheaper scores at least as well DeepSeek V3 0324
General intelligence: 15.4
$0.27/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)
General intelligence: 15.1
$0.10/M input
Dominated: something cheaper scores at least as well gpt-oss-20b (high)
General intelligence: 14.9
$0.07/M input
Dominated: something cheaper scores at least as well Nemotron 3 Nano Omni 30B A3B Reasoning
General intelligence: 14.9
$0.07/M input
Dominated: something cheaper scores at least as well gpt-oss-120b (low)
General intelligence: 14.9
$0.15/M input
Dominated: something cheaper scores at least as well GPT-4.1 mini
General intelligence: 14.8
$0.40/M input
Dominated: something cheaper scores at least as well Mistral Small 3.1
General intelligence: 14.7
$0.10/M input
Dominated: something cheaper scores at least as well Mistral Medium 3.1
General intelligence: 14.7
$0.40/M input
Dominated: something cheaper scores at least as well Qwen3 30B A3B 2507 (Reasoning)
General intelligence: 14.4
$0.20/M input
Dominated: something cheaper scores at least as well Nova 2.0 Pro Preview (Non-reasoning)
General intelligence: 14.4
$1/M input
Dominated: something cheaper scores at least as well gpt-oss-20b (low)
General intelligence: 14.3
$0.07/M input
Dominated: something cheaper scores at least as well GPT-5 mini (minimal)
General intelligence: 14.3
$0.25/M input
Dominated: something cheaper scores at least as well Llama 4 Maverick
General intelligence: 14.3
$0.27/M input
Dominated: something cheaper scores at least as well Qwen3 VL 235B A22B Instruct
General intelligence: 14.3
$0.70/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)
General intelligence: 14.2
$0.05/M input
Dominated: something cheaper scores at least as well DeepSeek V3 (Dec '24)
General intelligence: 14.2
$0.36/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash (Non-reasoning)
General intelligence: 14.1
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3 Next 80B A3B Instruct
General intelligence: 13.7
$0.50/M input
Dominated: something cheaper scores at least as well Qwen3 Coder 30B A3B Instruct
General intelligence: 13.6
$0.45/M input
Dominated: something cheaper scores at least as well QwQ 32B
General intelligence: 13.4
$0.66/M input
Dominated: something cheaper scores at least as well Qwen3 235B A22B (Reasoning)
General intelligence: 13.4
$0.70/M input
Dominated: something cheaper scores at least as well Qwen3 VL 30B A3B (Reasoning)
General intelligence: 13.3
$0.20/M input
Dominated: something cheaper scores at least as well Gemma 4 12B (Non-reasoning)
General intelligence: 13.2
$0.10/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)
General intelligence: 13.1
$0.10/M input
Dominated: something cheaper scores at least as well Nova Premier
General intelligence: 12.7
$3/M input
Dominated: something cheaper scores at least as well Mistral Medium 3
General intelligence: 12.5
$0.40/M input
Dominated: something cheaper scores at least as well Mistral Small 4 (Non-reasoning)
General intelligence: 12.4
$0.15/M input
Dominated: something cheaper scores at least as well Llama Nemotron Super 49B v1.5 (Reasoning)
General intelligence: 12.4
$0.40/M input
Dominated: something cheaper scores at least as well Nova 2.0 Lite (Non-reasoning)
General intelligence: 11.8
$0.30/M input
Dominated: something cheaper scores at least as well Claude 3 Opus
General intelligence: 11.8
$15/M input
Dominated: something cheaper scores at least as well Qwen3 32B (Reasoning)
General intelligence: 11.5
$0.70/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash-Lite (Reasoning)
General intelligence: 11.4
$0.10/M input
Dominated: something cheaper scores at least as well Magistral Small 1.2
General intelligence: 11.3
$0.50/M input
Dominated: something cheaper scores at least as well GPT-4o (Nov '24)
General intelligence: 11.2
$3/M input
Dominated: something cheaper scores at least as well Ministral 3 14B
General intelligence: 11.1
$0.20/M input
Dominated: something cheaper scores at least as well Qwen3 VL 32B Instruct
General intelligence: 11.1
$0.70/M input
Dominated: something cheaper scores at least as well GLM-4.6V (Non-reasoning)
General intelligence: 11
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3 235B A22B (Non-reasoning)
General intelligence: 10.9
$0.70/M input
Dominated: something cheaper scores at least as well Mistral Small 3.2
General intelligence: 10.6
$0.10/M input
Dominated: something cheaper scores at least as well Qwen3 VL 8B (Reasoning)
General intelligence: 10.6
$0.18/M input
Dominated: something cheaper scores at least as well Nova 2.0 Omni (Non-reasoning)
General intelligence: 10.5
$0.30/M input
Dominated: something cheaper scores at least as well Qwen3 14B (Reasoning)
General intelligence: 10.4
$0.35/M input
Dominated: something cheaper scores at least as well Llama 4 Scout
General intelligence: 10
$0.18/M input
Dominated: something cheaper scores at least as well Qwen3 VL 30B A3B Instruct
General intelligence: 10
$0.20/M input
Dominated: something cheaper scores at least as well DeepSeek R1 Distill Llama 70B
General intelligence: 9.9
$0.70/M input
Dominated: something cheaper scores at least as well Claude 3.5 Sonnet (Oct '24)
General intelligence: 9.9
$3/M input
Dominated: something cheaper scores at least as well GPT-4.1 nano
General intelligence: 9.6
$0.10/M input
Dominated: something cheaper scores at least as well Qwen3 Omni 30B A3B (Reasoning)
General intelligence: 9.6
$0.25/M input
Dominated: something cheaper scores at least as well Qwen2.5 Instruct 72B
General intelligence: 9.6
$0.47/M input
Dominated: something cheaper scores at least as well GPT-4o (Aug '24)
General intelligence: 9.6
$3/M input
Dominated: something cheaper scores at least as well Llama 3.3 Instruct 70B
General intelligence: 9.4
$0.60/M input
Dominated: something cheaper scores at least as well Qwen3 30B A3B (Reasoning)
General intelligence: 9.3
$0.20/M input
Dominated: something cheaper scores at least as well Qwen3 30B A3B 2507 Instruct
General intelligence: 9.1
$0.20/M input
Dominated: something cheaper scores at least as well GLM-4.5V (Reasoning)
General intelligence: 9.1
$0.60/M input
Dominated: something cheaper scores at least as well Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)
General intelligence: 9.1
$0.60/M input
Dominated: something cheaper scores at least as well Ministral 3 8B
General intelligence: 9
$0.15/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron Nano 12B v2 VL (Reasoning)
General intelligence: 9
$0.20/M input
Dominated: something cheaper scores at least as well Gemma 4 E4B (Non-reasoning)
General intelligence: 8.9
$0.02/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron Nano 9B V2 (Reasoning)
General intelligence: 8.8
$0.04/M input
Dominated: something cheaper scores at least as well Llama Nemotron Super 49B v1.5 (Non-reasoning)
General intelligence: 8.7
$0.40/M input
Dominated: something cheaper scores at least as well Qwen3 32B (Non-reasoning)
General intelligence: 8.6
$0.70/M input
Dominated: something cheaper scores at least as well GPT-4o (May '24)
General intelligence: 8.6
$5/M input
Dominated: something cheaper scores at least as well Llama 3.1 Instruct 405B
General intelligence: 8.5
$3/M input
Dominated: something cheaper scores at least as well Qwen3 VL 8B Instruct
General intelligence: 8.4
$0.18/M input
Dominated: something cheaper scores at least as well Qwen3 8B (Reasoning)
General intelligence: 8.3
$0.18/M input
Dominated: something cheaper scores at least as well Claude 3.5 Sonnet (June '24)
General intelligence: 8.3
$3/M input
Dominated: something cheaper scores at least as well GPT-5 nano (minimal)
General intelligence: 8
$0.05/M input
Dominated: something cheaper scores at least as well GPT-4 Turbo
General intelligence: 7.9
$10/M input
Dominated: something cheaper scores at least as well Nova Pro
General intelligence: 7.7
$0.80/M input
Dominated: something cheaper scores at least as well Command A
General intelligence: 7.7
$3/M input
Dominated: something cheaper scores at least as well Llama 3.1 Instruct 8B
General intelligence: 7.6
$0.08/M input
Dominated: something cheaper scores at least as well Llama 3.1 Nemotron Instruct 70B
General intelligence: 7.6
$1/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)
General intelligence: 7.4
$0.05/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron Nano 9B V2 (Non-reasoning)
General intelligence: 7.4
$0.05/M input
Dominated: something cheaper scores at least as well Mistral Large 2 (Jul '24)
General intelligence: 7.3
$2/M input
Dominated: something cheaper scores at least as well Qwen3 14B (Non-reasoning)
General intelligence: 7
$0.35/M input
Dominated: something cheaper scores at least as well GLM-4.5V (Non-reasoning)
General intelligence: 7
$0.60/M input
Dominated: something cheaper scores at least as well GPT-4
General intelligence: 7
$30/M input
Dominated: something cheaper scores at least as well Nova Lite
General intelligence: 6.9
$0.06/M input
Dominated: something cheaper scores at least as well Gemini 2.5 Flash-Lite (Non-reasoning)
General intelligence: 6.9
$0.10/M input
Dominated: something cheaper scores at least as well Mistral Small 3
General intelligence: 6.9
$0.10/M input
Dominated: something cheaper scores at least as well GPT-4o mini
General intelligence: 6.9
$0.15/M input
Dominated: something cheaper scores at least as well Ministral 3 3B
General intelligence: 6.8
$0.10/M input
Dominated: something cheaper scores at least as well Qwen3 30B A3B (Non-reasoning)
General intelligence: 6.8
$0.20/M input
Dominated: something cheaper scores at least as well Llama 3.1 Instruct 70B
General intelligence: 6.8
$0.56/M input
Dominated: something cheaper scores at least as well Granite 4.1 8B
General intelligence: 6.7
$0.05/M input
Dominated: something cheaper scores at least as well Qwen2.5 Turbo
General intelligence: 6.3
$0.05/M input
Dominated: something cheaper scores at least as well Jamba 1.7 Large
General intelligence: 5.3
$2/M input
Dominated: something cheaper scores at least as well Granite 4.0 H Small
General intelligence: 5.2
$0.06/M input
Dominated: something cheaper scores at least as well Qwen3 8B (Non-reasoning)
General intelligence: 5.1
$0.18/M input
Dominated: something cheaper scores at least as well Qwen3 Omni 30B A3B Instruct
General intelligence: 5.1
$0.25/M input
Dominated: something cheaper scores at least as well Jamba 1.5 Large
General intelligence: 5.1
$2/M input
Dominated: something cheaper scores at least as well Jamba 1.6 Large
General intelligence: 5
$2/M input
Dominated: something cheaper scores at least as well Phi-4
General intelligence: 4.9
$0.13/M input
Dominated: something cheaper scores at least as well Nova Micro
General intelligence: 4.7
$0.04/M input
Dominated: something cheaper scores at least as well Mistral Small (Sep '24)
General intelligence: 4.7
$0.20/M input
Dominated: something cheaper scores at least as well Claude 3 Sonnet
General intelligence: 4.7
$3/M input
Dominated: something cheaper scores at least as well NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)
General intelligence: 4.6
$0.20/M input
Dominated: something cheaper scores at least as well Mistral Large (Feb '24)
General intelligence: 4.4
$4/M input
Dominated: something cheaper scores at least as well Llama 2 Chat 7B
General intelligence: 4.3
$0.05/M input
Dominated: something cheaper scores at least as well Claude 3 Haiku
General intelligence: 3.9
$0.25/M input
Dominated: something cheaper scores at least as well Mistral Small (Feb '24)
General intelligence: 3.6
$0.15/M input
Dominated: something cheaper scores at least as well GPT-3.5 Turbo
General intelligence: 3.6
$0.50/M input
Dominated: something cheaper scores at least as well Mistral Medium
General intelligence: 3.6
$2/M input
Dominated: something cheaper scores at least as well Llama 3 Instruct 70B
General intelligence: 3.5
$0.65/M input
Dominated: something cheaper scores at least as well Llama 3.2 Instruct 11B (Vision)
General intelligence: 3.3
$0.34/M input
Dominated: something cheaper scores at least as well Command-R+ (Apr '24)
General intelligence: 3
$3/M input
Dominated: something cheaper scores at least as well Jamba 1.5 Mini
General intelligence: 2.7
$0.20/M input
Dominated: something cheaper scores at least as well Jamba 1.6 Mini
General intelligence: 2.6
$0.20/M input
Dominated: something cheaper scores at least as well Mixtral 8x7B Instruct
General intelligence: 2.4
$0.45/M input
Dominated: something cheaper scores at least as well Mistral 7B Instruct
General intelligence: 2.1
$0.25/M input
Dominated: something cheaper scores at least as well Command-R (Mar '24)
General intelligence: 2.1
$0.50/M input
Dominated: something cheaper scores at least as well Granite 3.3 8B (Non-reasoning)
General intelligence: 1.8
$0.03/M input
Dominated: something cheaper scores at least as well Llama 3 Instruct 8B
General intelligence: 1.2
$0.04/M input
Dominated: something cheaper scores at least as well Gemma 3n E4B Instruct
General intelligence: 1.2
$0.06/M input
Dominated: something cheaper scores at least as well Claude Opus 5 (Adaptive Reasoning, Max Effort)
General intelligence: 60.7
$5/M input
On the frontier: no cheaper model scores as well Kimi K3
General intelligence: 57.1
$3/M input
On the frontier: no cheaper model scores as well GPT-5.6 Terra (max)
General intelligence: 55
$3/M input
On the frontier: no cheaper model scores as well Grok 4.5 (high)
General intelligence: 53.8
$2/M input
On the frontier: no cheaper model scores as well GPT-5.6 Luna (max)
General intelligence: 51.2
$0.20/M input
On the frontier: no cheaper model scores as well DeepSeek V4 Flash (Reasoning, Max Effort)
General intelligence: 40.3
$0.14/M input
On the frontier: no cheaper model scores as well Gemma 4 26B A4B (Reasoning)
General intelligence: 25.7
$0.13/M input
On the frontier: no cheaper model scores as well GLM-4.7-Flash (Reasoning)
General intelligence: 22.9
$0.07/M input
On the frontier: no cheaper model scores as well Qwen3.5 4B (Reasoning)
General intelligence: 20.1
$0.03/M input
On the frontier: no cheaper model scores as well Gemma 4 E4B (Reasoning)
General intelligence: 11.9
$0.02/M input
On the frontier: no cheaper model scores as well Gemma 4 E4B Qwen3.5 4B Claude Opus 5 Intelligence Index v4.1 input price $/M, log
Read it like this. Claude Opus 5 scores 60.7 at $5/M and sits on the frontier: nothing cheaper matches it. Claude Fable 5 scores 59.9 at $10/M, which is a lower score at double the price, so it is dominated. On this chart that is one dot up and to the left of another, and it is the whole argument for checking price against capability before signing anything.
How this is derived