{
 "sym": "Fm",
 "updated": "2026-08-06",
 "verdict": "Claude, and in Aug 2026 it is no longer close on the measures a founder should care about: Anthropic holds the #1 Artificial Analysis Intelligence Index score (Opus 5, 61) plus three of the top five slots, 40% of enterprise LLM API spend (54% in coding), and a $47B revenue run-rate that passed OpenAI's in May 2026. ChatGPT/GPT-5.6 is the pick when consumer distribution, ecosystem breadth, or cheap volume decide \u2014 Luna at $0.20/$1.20 per MTok is the lowest price any frontier lab has ever printed. Gemini wins when you live on Google surfaces, need 1M-context multimodal at mid-tier prices, or want the strongest bundled-subscription value. Grok when real-time X data or the cheapest frontier-flagship API matters. Meta AI when free frontier-class answers beat having an API at all.",
 "top5": [
  {
   "rank": 1,
   "name": "Claude",
   "maker": "Anthropic",
   "url": "https://claude.com",
   "docs": "https://platform.claude.com/docs",
   "pricing": "Free \u00b7 Pro $17\u201320/mo \u00b7 Max from $100/mo \u00b7 Team $20\u201325/seat \u00b7 API: Opus 5 $5/$25, Fable 5 $10/$50 per MTok",
   "best_for": "The primary model a founder goes deep on \u2014 hardest reasoning, long-horizon agentic work, and the whole Claude Code/Cowork surface on one subscription.",
   "why": "Owns the top of every serious capability index: Opus 5 leads the Artificial Analysis Intelligence Index at 61 and Anthropic holds three of the top five slots (Aug 2026); Fable 5 (Jun 9, 2026) was the first model past 90% on core analytics benchmarks with a 1M-token window. The market agrees \u2014 40% of enterprise LLM API spend (54% in coding, Menlo Ventures) and a $47B run-rate by May 2026, up from $9B at end-2025, now the largest of any AI lab.",
   "watch": "Consumer distribution is tiny \u2014 ~2.1% of chatbot web traffic vs ChatGPT's ~57\u201367%. Fable 5 shipped, was suspended three days later (Jun 12), and only redeployed Jul 1 \u2014 frontier-tier access can move under you. Fable 5's $10/$50 pricing is the most expensive mainstream API on the market.",
   "evidence": [
    {
     "stat": "#1 Artificial Analysis Intelligence Index: Claude Opus 5 at 61; Anthropic holds 3 of the top 5 slots (Aug 2026)",
     "src": "https://artificialanalysis.ai/models"
    },
    {
     "stat": "$47B run-rate revenue (May 2026, from $9B end-2025); $65B Series H at $965B post-money (May 28, 2026)",
     "src": "https://www.anthropic.com/news/series-h"
    },
    {
     "stat": "40% of enterprise LLM API spend, 54% in coding (Menlo Ventures report, 2025\u21922026)",
     "src": "https://officechai.com/ai/anthropic-google-gain-market-share-at-openais-expense-in-enterprise-api-llm-usage-menlo-ventures-report/"
    }
   ],
   "tile_note": "deep reasoning, enterprise leader"
  },
  {
   "rank": 2,
   "name": "ChatGPT / GPT-5.6",
   "maker": "OpenAI",
   "url": "https://openai.com",
   "docs": "https://platform.openai.com/docs",
   "pricing": "Free \u00b7 Go ~$4.50/mo \u00b7 Plus $20/mo \u00b7 Pro $200/mo \u00b7 Team $25/seat \u00b7 API: Sol $5/$30, Terra $2/$12, Luna $0.20/$1.20 per MTok (after Jul 30 cuts)",
   "best_for": "Broadest ecosystem and cheapest high-quality volume \u2014 the default when your users, integrations, and hires already live in ChatGPT.",
   "why": "GPT-5.6 (Jul 9, 2026) split the flagship into durable Sol/Terra/Luna tiers, and the Jul 30 price cuts made Luna ($0.20/$1.20) the cheapest near-frontier API ever sold. Distribution is unmatched: ~900M weekly ChatGPT users (Feb 2026), 50M paying subscribers, $25B annualized revenue \u2014 the largest consumer AI footprint on earth.",
   "watch": "The enterprise story is going the other way \u2014 API share fell from 50% (2023) to 27% (Menlo), and Sol (59) sits below Opus 5 (61) on the Intelligence Index. Still deeply unprofitable (~$14B/yr losses reported) despite a record $122B raise at $852B (Mar 2026). Pricing has four compounding multipliers (fast mode, long-context 2x over 272k, caching, residency) \u2014 worst-case Sol bills run ~4x list.",
   "evidence": [
    {
     "stat": "~900M weekly active users (Feb 27, 2026); $25B annualized revenue (Mar 2026); 50M paid subscribers",
     "src": "https://technologychecker.io/blog/chatgpt-statistics"
    },
    {
     "stat": "GPT-5.6 launch Jul 9, 2026; Luna cut to $0.20/$1.20, Terra to $2/$12 per MTok on Jul 30, 2026",
     "src": "https://www.eesel.ai/blog/gpt-5-6-pricing"
    },
    {
     "stat": "Enterprise API share 50% (2023) \u2192 27% (Menlo Ventures, 2025\u21922026 report)",
     "src": "https://officechai.com/ai/anthropic-google-gain-market-share-at-openais-expense-in-enterprise-api-llm-usage-menlo-ventures-report/"
    }
   ],
   "tile_note": "the ecosystem default"
  },
  {
   "rank": 3,
   "name": "Gemini",
   "maker": "Google (DeepMind)",
   "url": "https://gemini.google",
   "docs": "https://ai.google.dev",
   "pricing": "Free \u00b7 AI Plus $4.99/mo \u00b7 AI Pro $19.99/mo \u00b7 AI Ultra $99.99 (5x) / $199.99 (20x) \u00b7 API: 3.1 Pro $2/$12 per MTok (\u2264200k ctx; higher above)",
   "best_for": "Multimodal work, 1M-token context at mid-tier prices, and founders already living on Google Workspace/Cloud \u2014 the best bundled-subscription value in AI.",
   "why": "Gemini 3.1 Pro (GA Feb 19, 2026) is the abstract-reasoning standout \u2014 94.3% GPQA Diamond, 77.1% ARC-AGI-2, 80.6% SWE-bench Verified \u2014 at $2/$12, a third of Sol's output price. Distribution is compounding: 900M Gemini app MAU (doubled YoY) and 1B+ AI Mode users announced at I/O 2026, with the subscription bundling TB-scale storage, Flow credits, and the Jules coding agent.",
   "watch": "The flagship succession slipped: Gemini 3.5 Pro was announced May 19 at I/O with a 'next month' target and still hadn't shipped by early Aug \u2014 only 3.5 Flash made it out. Current 3.1 Pro sits below the Anthropic/OpenAI leaders on the Intelligence Index, and the Plus/Pro/Ultra\u00d7Flow-credits tiering is genuinely confusing to buy.",
   "evidence": [
    {
     "stat": "Gemini app 900M MAU, 2x YoY; Search AI Mode past 1B MAU (Google I/O, May 2026)",
     "src": "https://www.startuphub.ai/ai-news/ai-figures/2026/figure-sundar-pichai-google-io-2026-keynote-recap-2026-06-01"
    },
    {
     "stat": "Gemini 3.1 Pro GA Feb 19, 2026: GPQA 94.3%, ARC-AGI-2 77.1%, SWE-bench Verified 80.6%, 1M ctx at $2/$12",
     "src": "https://theairankings.com/google/gemini-3-1-pro/"
    },
    {
     "stat": "Gemini 3.5 Pro announced May 19, 2026, still unreleased as of Jul 2026 \u2014 only 3.5 Flash shipped",
     "src": "https://www.eesel.ai/blog/gemini-3-5-pro"
    }
   ],
   "tile_note": "multimodal + google surface"
  },
  {
   "rank": 4,
   "name": "Grok",
   "maker": "xAI (SpaceX, SPCX)",
   "url": "https://x.ai",
   "docs": "https://docs.x.ai",
   "pricing": "Free \u00b7 X Premium $8/mo \u00b7 SuperGrok $30/mo \u00b7 SuperGrok Plus $100 \u00b7 Heavy $300 \u00b7 API: Grok 4.5 $2/$6, Grok 4.3 $1.25/$2.50 per MTok",
   "best_for": "Real-time X/web data baked into the model, native video generation, and the cheapest frontier-flagship API on the market.",
   "why": "Grok 4.5 (Jul 8, 2026) prices a frontier flagship at $2/$6 with a 500k context \u2014 a fraction of Fable 5 or Sol \u2014 and is the only frontier model with live X firehose grounding plus native 2K video generation. Backed by the Colossus buildout (~555k GPUs, ~2GW) and now inside the public SpaceX entity after the Feb 2, 2026 all-stock absorption and Jun 12 SPCX IPO.",
   "watch": "Commercially it is a distant fourth: ~$500M annualized run-rate mid-2026 vs Anthropic's $47B and OpenAI's $25B, and 117M MAU vs Gemini's 900M. June 2026 moved all paid tiers to one shared weekly usage pool and retired budget models with two months' notice \u2014 plan churn is real. Moderation controversies keep it off many enterprise shortlists.",
   "evidence": [
    {
     "stat": "Grok 4.5 flagship (Jul 8, 2026): $2/$6 per MTok, 500k context; SuperGrok $30/mo, Heavy $300/mo",
     "src": "https://felloai.com/grok-pricing/"
    },
    {
     "stat": "xAI absorbed into SpaceX Feb 2, 2026 ($230B standalone); SPCX IPO Jun 12, 2026, ~$1.49T market cap by late Jul",
     "src": "https://valueaddvc.com/blog/xai-valuation-2026-the-230b-series-e-the-125t-spacex-merger-and-what-grok-is-worth-today"
    },
    {
     "stat": "Grok MAU 64M (early 2026) \u2192 117M (Mar 2026); ~$500M annualized revenue mid-2026",
     "src": "https://valueaddvc.com/blog/xai-valuation-2026-the-230b-series-e-the-125t-spacex-merger-and-what-grok-is-worth-today"
    }
   ],
   "tile_note": "real-time x, cheapest flagship"
  },
  {
   "rank": 5,
   "name": "Meta AI (Muse Spark)",
   "maker": "Meta Superintelligence Labs",
   "url": "https://meta.ai",
   "docs": "https://ai.meta.com",
   "pricing": "Free on meta.ai and Meta AI app \u00b7 API in private preview, no public pricing",
   "best_for": "Free frontier-class multimodal answers at consumer scale \u2014 and a bet that Meta's closed-model pivot produces a real fifth frontier seat.",
   "why": "Muse Spark (Apr 8, 2026) is Meta's first frontier model and its first without open weights \u2014 top-5 on the Artificial Analysis Intelligence Index at launch (52) and 50.2% on Humanity's Last Exam, with strong multimodal and visual-STEM performance. Distribution through Meta's app family gives it the largest free frontier audience anywhere.",
   "watch": "Not really buyable yet: the API is private-preview with no published pricing, and it trails rivals on coding and long-horizon agentic work (1,427 vs Claude Sonnet's 1,648 on agent tasks). Strategy whiplash \u2014 open-weights champion to closed model in one generation, with leadership saying future versions may reopen \u2014 makes it hard to build on. Enterprise share fell 16% \u2192 8% (Menlo).",
   "evidence": [
    {
     "stat": "Muse Spark launched Apr 8, 2026 \u2014 Meta's first closed-weights frontier model; Intelligence Index 52 (top-5 at launch), HLE 50.2%",
     "src": "https://the-decoder.com/metas-muse-spark-is-its-first-frontier-model-and-its-first-without-open-weights/"
    },
    {
     "stat": "Meta enterprise LLM API share 16% (2023) \u2192 8% (Menlo Ventures, 2025\u21922026)",
     "src": "https://officechai.com/ai/anthropic-google-gain-market-share-at-openais-expense-in-enterprise-api-llm-usage-menlo-ventures-report/"
    }
   ],
   "tile_note": "free frontier, no api"
  }
 ],
 "matrix": {
  "cols": [
   "Flagship (Aug 2026)",
   "API $/1M in\u00b7out",
   "Context",
   "Consumer entry",
   "Intelligence Index",
   "Ent. API share",
   "Edge"
  ],
  "rows": [
   [
    "Claude",
    "Fable 5 \u00b7 Opus 5",
    "$10/$50 \u00b7 $5/$25",
    "1M",
    "Pro $17\u201320/mo",
    "61 (#1, Opus 5)",
    "40% (54% coding)",
    "Agentic depth, Code/Cowork surface"
   ],
   [
    "ChatGPT / GPT-5.6",
    "GPT-5.6 Sol/Terra/Luna",
    "$5/$30 \u2192 $0.20/$1.20",
    "~1M (2x price >272k)",
    "Plus $20/mo",
    "59 (Sol max)",
    "27%",
    "Distribution, cheapest quality volume"
   ],
   [
    "Gemini",
    "3.1 Pro + 3.5 Flash",
    "$2/$12 (\u2264200k)",
    "1M",
    "AI Pro $19.99/mo",
    "below top 5 (3.5 Pro delayed)",
    "21%",
    "Multimodal, bundle value, ARC-AGI"
   ],
   [
    "Grok",
    "Grok 4.5",
    "$2/$6",
    "500k",
    "SuperGrok $30/mo",
    "not top 5",
    "low single digits",
    "Real-time X, native video, price"
   ],
   [
    "Meta AI (Muse Spark)",
    "Muse Spark (closed)",
    "private preview",
    "n/d",
    "free",
    "52 (top-5 at Apr launch)",
    "8% (Llama)",
    "Free at consumer scale"
   ],
   [
    "Mistral (\u2192Ow)",
    "Medium 3.5 / Large 3 (open)",
    "low, open-weight",
    "up to 1M",
    "Le Chat/Vibe free\u2013paid",
    "trails frontier",
    "~1\u20132%",
    "EU sovereignty, Apache weights"
   ],
   [
    "DeepSeek (\u2192Ow)",
    "V4 / V4-Flash",
    "$0.14/$0.28 (Flash)",
    "1M-class",
    "free app",
    "50 (V4-Flash)",
    "~1%",
    "The price floor"
   ]
  ]
 },
 "rules": [
  {
   "if": "You're picking your one primary model \u2014 the brief's own advice \u2014 and your work is judgment-heavy",
   "then": "Claude. Start on Opus 5 ($5/$25, Intelligence Index #1) and escalate to Fable 5 only for the hardest long-horizon work; a Pro/Max subscription carries Claude Code and Cowork with it."
  },
  {
   "if": "You're shipping AI inside a consumer product, or burning tokens at volume on routine tasks",
   "then": "GPT-5.6 Luna/Terra \u2014 $0.20/$1.20 and $2/$12 after the Jul 30 cuts undercut every frontier rival; keep Sol for the hard 5% of calls."
  },
  {
   "if": "Your workload is multimodal (video, audio, images, PDFs) or lives on Google Workspace/Cloud",
   "then": "Gemini 3.1 Pro at $2/$12 with 1M context, and the AI Pro bundle at $19.99 \u2014 the strongest per-dollar consumer subscription in AI."
  },
  {
   "if": "You need real-time social/news signal or the cheapest frontier-flagship API",
   "then": "Grok 4.5 at $2/$6 \u2014 accepting the plan-churn and the fact that xAI's commercial traction is a rounding error next to the top three."
  },
  {
   "if": "Your bill, privacy, or latency argument beats the capability argument",
   "then": "Don't buy a bigger frontier subscription \u2014 go to element Ow: DeepSeek V4-Flash (Index 50 at $0.14/$0.28) and Kimi K3 (Index 57, open) have closed most of the gap at ~1/10 the price."
  }
 ],
 "field": [
  {
   "name": "Claude Mythos 5",
   "maker": "Anthropic",
   "note": "Fable 5 without safety classifiers \u2014 restricted to Project Glasswing partners and vetted researchers (cyber, bio); same $10/$50 pricing",
   "url": "https://platform.claude.com/docs/en/about-claude/models/introducing-claude-fable-5-and-claude-mythos-5",
   "oss": false,
   "entry": "approved customers only",
   "status": "active"
  },
  {
   "name": "Microsoft MAI",
   "maker": "Microsoft AI",
   "note": "Seven in-house models at Build 2026 (MAI-Thinking-1, Code, Voice, Image); competitive mid-weight, not frontier \u2014 distributed via Foundry, OpenRouter, Copilot",
   "url": "https://microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models/",
   "oss": false,
   "entry": "via Azure Foundry / Copilot",
   "status": "active"
  },
  {
   "name": "Amazon Nova 2",
   "maker": "AWS",
   "note": "Value-tier model family on Bedrock \u2014 50k+ customers, priced against Haiku-class, not a frontier contender",
   "url": "https://aws.amazon.com/ai/generative-ai/nova/",
   "oss": false,
   "entry": "Bedrock usage-based",
   "status": "active"
  },
  {
   "name": "Gemini Spark",
   "maker": "Google",
   "note": "24/7 cloud agentic assistant for AI Ultra subscribers \u2014 the subscription hook for Ultra tiers",
   "url": "https://gemini.google/subscriptions/",
   "oss": false,
   "entry": "AI Ultra $99.99+/mo",
   "status": "active"
  },
  {
   "name": "DeepSeek V4",
   "maker": "DeepSeek",
   "note": "Frontier-class open-weight value leader (V4-Flash: Index 50 at $0.14/$0.28); first external round at $52\u201359B (2026) \u2014 primary home is element Ow",
   "url": "https://www.deepseek.com",
   "oss": true,
   "entry": "free app + cheap API",
   "status": "active"
  },
  {
   "name": "Kimi K3",
   "maker": "Moonshot AI",
   "note": "2.8T-param open-weight flagship (Jul 2026), Intelligence Index 57 \u2014 #3 overall; HK IPO targeted Q4 2026 \u2014 belongs to element Ow",
   "url": "https://kimi.com",
   "oss": true,
   "entry": "free + tokens",
   "status": "active"
  },
  {
   "name": "Mistral / Le Chat (Vibe)",
   "maker": "Mistral AI",
   "note": "EU-sovereign lab, ~$400M ARR (Jan 2026), \u20ac3B raise talks at ~\u20ac20B; flagships ship open-weight, so primary home is element Ow",
   "url": "https://mistral.ai",
   "oss": true,
   "entry": "free \u00b7 paid Le Chat tiers",
   "status": "active"
  },
  {
   "name": "GLM-5.2",
   "maker": "Z.ai (Zhipu)",
   "note": "Open-weight frontier chaser from the Chinese wave matching proprietary rivals on several boards \u2014 element Ow",
   "url": "https://z.ai",
   "oss": true,
   "entry": "free + tokens",
   "status": "active"
  },
  {
   "name": "Qwen",
   "maker": "Alibaba",
   "note": "Broad open-weight family plus hosted apps; a top OpenRouter presence \u2014 element Ow",
   "url": "https://qwen.ai",
   "oss": true,
   "entry": "free + tokens",
   "status": "active"
  },
  {
   "name": "Llama frontier line",
   "maker": "Meta",
   "note": "Llama 4 Scout/Maverick (Apr 2025) were the last open flagships; the Behemoth-class successor never shipped open \u2014 Meta's frontier effort moved to closed Muse Spark",
   "url": "https://www.llama.com",
   "oss": true,
   "entry": "free weights",
   "status": "fading"
  },
  {
   "name": "Cohere",
   "maker": "Cohere",
   "note": "Exited the frontier race; pivoted to private-deployment enterprise AI (North platform) \u2014 no longer a general-model contender",
   "url": "https://cohere.com",
   "oss": false,
   "entry": "enterprise contracts",
   "status": "fading"
  },
  {
   "name": "AI21 Labs",
   "maker": "AI21",
   "note": "Jamba long-context line continues but far from frontier pace",
   "url": "https://www.ai21.com",
   "oss": true,
   "entry": "free + tokens",
   "status": "fading"
  },
  {
   "name": "Reka",
   "maker": "Reka AI",
   "note": "Small multimodal lab; capable mid-tier models, no frontier presence in 2026",
   "url": "https://reka.ai",
   "oss": false,
   "entry": "usage-based",
   "status": "fading"
  },
  {
   "name": "Inflection AI (Pi)",
   "maker": "Inflection",
   "note": "The original consumer-companion frontier bid; core team absorbed by Microsoft (2024), pivoted to enterprise \u2014 Pi effectively frozen",
   "url": "https://inflection.ai",
   "oss": false,
   "entry": "\u2014",
   "status": "dead"
  },
  {
   "name": "xAI (standalone)",
   "maker": "xAI \u2192 SpaceX",
   "note": "Ceased to exist as an independent company Feb 2, 2026 \u2014 absorbed into SpaceX at $230B, now part of public SPCX",
   "url": "https://x.ai",
   "oss": false,
   "entry": "\u2014",
   "status": "acquired"
  },
  {
   "name": "Grok 3 / 4.1 Fast",
   "maker": "xAI",
   "note": "Budget model line retired May 15, 2026 with ~2 months' notice; requests redirect to Grok 4.3 at its rates",
   "url": "https://felloai.com/grok-pricing/",
   "oss": false,
   "entry": "\u2014",
   "status": "dead"
  }
 ],
 "signals": [
  {
   "fact": "Enterprise LLM API spend hit $8.4B by mid-2025 (from $3.5B six months prior) and kept doubling; share now Anthropic 40% / OpenAI 27% / Google 21% / Meta 8% (Menlo Ventures)",
   "src": "https://officechai.com/ai/anthropic-google-gain-market-share-at-openais-expense-in-enterprise-api-llm-usage-menlo-ventures-report/"
  },
  {
   "fact": "Anthropic passed OpenAI in revenue: $47B run-rate (May 2026) vs OpenAI ~$25\u201333B; $65B Series H at $965B post and a confidential IPO filing (Jun 1, 2026)",
   "src": "https://www.anthropic.com/news/series-h"
  },
  {
   "fact": "OpenAI raised a record $122B at $852B post (Mar 2026) while running ~$14B/yr losses \u2014 scale-at-all-costs is still the strategy",
   "src": "https://cheatsheets.davidveksler.com/ai-frontier.html"
  },
  {
   "fact": "xAI merged into SpaceX (Feb 2, 2026); combined SPCX IPO'd Jun 12, 2026 at $135/share \u2014 the first public frontier lab, ~$1.49T market cap by late Jul",
   "src": "https://valueaddvc.com/blog/xai-valuation-2026-the-230b-series-e-the-125t-spacex-merger-and-what-grok-is-worth-today"
  },
  {
   "fact": "Frontier price war: OpenAI cut Luna to $0.20/$1.20 and Terra to $2/$12 (Jul 30, 2026) as open-weight rivals (Kimi K3 Index 57, DeepSeek V4-Flash Index 50 at $0.14/$0.28) closed the capability gap",
   "src": "https://www.eesel.ai/blog/gpt-5-6-pricing"
  },
  {
   "fact": "The open-weights era ended at the top: Meta shipped its first closed frontier model (Muse Spark, Apr 8, 2026), completing the 'great re-closing' of US frontier labs",
   "src": "https://the-decoder.com/metas-muse-spark-is-its-first-frontier-model-and-its-first-without-open-weights/"
  }
 ],
 "notes": "Charter: ranked on verified capability (Artificial Analysis Intelligence Index, arena boards), what founders actually pay (vendor pricing pages, all checked Aug 6, 2026), enterprise adoption (Menlo Ventures), and financial durability \u2014 no affiliate consideration. Scope: proprietary frontier subscriptions/APIs only \u2014 open-weight families (Llama, DeepSeek, Kimi, GLM, Qwen, Mistral's flagships) \u2192 Ow; routers (OpenRouter, LiteLLM) \u2192 Rt; coding agents built on these models \u2192 Ca; Perplexity and deep-research products \u2192 Re. Conflicts resolved: (1) Menlo enterprise share \u2014 we cite the later 2025\u21922026 report (Anthropic 40/OpenAI 27) over the mid-2025 cut (32/25); (2) Google AI Ultra pricing \u2014 vendor page ($99.99 5x / $199.99 20x) over I/O-recap paraphrases ('$250\u2192$200'); (3) Claude Pro $17\u201320 and Team $20\u201325 from claude.com/pricing over Sacra's $20/$30; (4) OpenAI revenue \u2014 $25B annualized (Mar 2026) is the consistently reported figure, Sacra's '$33B by May' is higher and unconfirmed, flagged accordingly; (5) benchmark leader \u2014 Intelligence Index (Opus 5 #1) and LMArena text (Fable 5 ~1508) currently agree on Anthropic, but Gemini 3.1 Pro leads ARC-AGI-style abstract reasoning; we weight the composite index. Weak-source flags: OpenAI 900M-WAU/$25B and Grok MAU/revenue figures come from stat aggregators tracing to press reports, not filings; the $122B OpenAI round and Meta's agent-task Elo are effectively single-source; Fable 5's HLE 53.3% is aggregator-only. Meta's Muse Spark API remains private-preview with no public pricing \u2014 its #5 rank is a distribution-and-trajectory call, not a buyability one. Gemini 3.5 Pro was announced May 19, 2026 and had still not shipped at press time; treat any '3.5 Pro benchmark' claims as vapor until GA.",
 "sources": [
  "https://artificialanalysis.ai/models",
  "https://www.anthropic.com/news/claude-fable-5-mythos-5",
  "https://www.anthropic.com/news/claude-opus-5",
  "https://www.anthropic.com/news/series-h",
  "https://platform.claude.com/docs/en/about-claude/models/introducing-claude-fable-5-and-claude-mythos-5",
  "https://claude.com/pricing",
  "https://openai.com/index/gpt-5-6/",
  "https://www.eesel.ai/blog/gpt-5-6-pricing",
  "https://www.cnbc.com/2026/07/30/open-ai-price-cut-gpt.html",
  "https://theairankings.com/google/gemini-3-1-pro/",
  "https://gemini.google/subscriptions/",
  "https://www.eesel.ai/blog/gemini-3-5-pro",
  "https://www.startuphub.ai/ai-news/ai-figures/2026/figure-sundar-pichai-google-io-2026-keynote-recap-2026-06-01",
  "https://felloai.com/grok-pricing/",
  "https://felloai.com/best-ai-models/",
  "https://valueaddvc.com/blog/xai-valuation-2026-the-230b-series-e-the-125t-spacex-merger-and-what-grok-is-worth-today",
  "https://the-decoder.com/metas-muse-spark-is-its-first-frontier-model-and-its-first-without-open-weights/",
  "https://officechai.com/ai/anthropic-google-gain-market-share-at-openais-expense-in-enterprise-api-llm-usage-menlo-ventures-report/",
  "https://konabayev.com/blog/llm-market-share/",
  "https://technologychecker.io/blog/chatgpt-statistics",
  "https://sacra.com/c/anthropic/",
  "https://cheatsheets.davidveksler.com/ai-frontier.html",
  "https://microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models/",
  "https://www.swfte.com/ai/leaderboard"
 ],
 "element": {
  "number": 1,
  "name": "Frontier Model",
  "group": "Intelligence",
  "essential": true,
  "edition": "v2026.Q3",
  "revision": "r7",
  "license": "CC BY 4.0 \u2014 cite elems.ai",
  "url": "https://elems.ai/e/fm.html"
 }
}