moatwatch
an mcgrathlabs observatoryTracking the commoditization of intelligence: how fast open-weight models close on the frontier, what a token actually costs — API, hosted open, or owned hardware — and pre-registered turning points so the regime change is visible before it’s a headline.
As of 2026-07-18
Turning points
Pre-registered conditions, evaluated daily. A trip is a signal, not a narrative.
Open weights reach the frontier
Best open-weight model comes within 5% of the best closed model on the Artificial Analysis Intelligence Index. Intelligence is commoditized one generation sooner than expected.
Trip condition: gapPct ≤ 5 (warming ≤ 10)
Open weights carry most real traffic
Open-weight models exceed 50% of OpenRouter daily token volume. Revealed preference — the market votes with paid tokens, not benchmarks.
Trip condition: openTokenShare ≥ 50 (warming ≥ 40) · tripped 2026-03-19
Owning beats renting at realistic utilization
A sub-$10k prosumer rig at 60% utilization produces tokens cheaper than the cheapest hosted near-frontier open model. The moment self-hosting stops being a fantasy for small teams.
Trip condition: prosumer $/Mtok @60% util ≤ hosted open floor (warming ≤ 1.5×)
Frontier price war goes vertical
Blended $ per intelligence-index point of the best closed model falls 50%+ within 90 days. The labs are torching margin to defend share — commoditization priced in by the producers themselves.
Trip condition: frontierPricePerIntel drops ≥50% over trailing 90d (warming ≥ 30%)
Hyperscalers and semis fall together
Both the hyperscaler basket and semis (SMH) sit 15%+ below their trailing 90-day highs at once. Not a rotation — the whole AI capex trade repricing simultaneously, the Cisco-2000 shape.
Trip condition: min(|drawdown|) of both baskets ≥ 15% (warming ≥ 10%)
Spenders and suppliers decouple
Hyperscaler basket and semis 60-day returns spread beyond 15 points. The see-saw scenario: the market rewarding capex cuts (or punishing them) instead of moving both sides together.
Trip condition: |hyperscaler 60d return − SMH 60d return| ≥ 15pp (warming ≥ 10)
Market splits frontier from commodity silicon
Nvidia's 60-day return runs 20+ points ahead of the commodity-silicon basket (AMD, AVGO, MRVL, MU, TSM). Open-weight commoditization pricing into the chip complex — inference silicon commoditizes while frontier-training silicon keeps its moat.
Trip condition: NVDA 60d return − commodity basket 60d return ≥ 20pp (warming ≥ 12)
A hyperscaler blinks on capex
Any top-5 hyperscaler guides capital expenditure down year-over-year. The single loudest 'something gave' signal available — tracked manually from earnings guidance in fundamentals.json.
Trip condition: guidance cuts YoY ≥ 1 (manual quarterly update)
The frontier gap
Best open-weight model vs best closed model, Artificial Analysis Intelligence Index (% behind). Shaded region is curated backfill; numbered lines are key releases.
- 1. GPT-4 — Frontier gap at its widest — no credible open competition.
- 2. Llama 2 — First serious open-weight release with a commercial license.
- 3. Llama 3.1 405B — First open model marketed as frontier-class.
- 4. DeepSeek R1 — The 'DeepSeek moment' — $600B single-day Nvidia drawdown.
- 5. Kimi K2 — 1T-param open MoE; agentic benchmarks near closed models.
- 6. Price-war reports — WSJ: OpenAI weighs deep price cuts to defend enterprise share vs Anthropic.
- 7. Kimi K3 — 2.8T open MoE within ~7% of frontier; SOX enters bear market territory.
The price of a token
Blended $/Mtok (3:1 input:output), log scale. Frontier list price vs the cheapest hosted near-frontier open model.
Where the tokens actually run
Open-weight share of OpenRouter daily token volume — revealed preference, not benchmarks.
Own vs rent
Effective $/Mtok by sourcing strategy at your volume. The turning point: when a sub-$10k box beats the hosted floor at realistic utilization.
Local cost = hardware amortized over 36 months ÷ monthly volume + electricity. Excludes ops labor, cooling, and redundancy — real self-hosting is worse than this, which makes a local win here a strong signal.
The capex reckoning: spenders vs suppliers
Equal-weight hyperscaler basket (AMZN MSFT GOOGL META ORCL) vs semis (SMH), indexed to 100 two years ago. Rising together = the boom; falling together = the correlated unwind.
Drawdown from trailing 90-day high
The unwind gauge: both baskets 15%+ underwater at once is the trip condition — a simultaneous repricing, not a rotation.
Frontier silicon vs commodity silicon
Nvidia vs equal-weight AMD/AVGO/MRVL/MU/TSM, indexed. When open models commoditize inference, the market splits these two — this week's Marvell-down-Nvidia-flat pattern, as a series.
The capex bill
Hand-curated from earnings guidance and analyst estimates, updated 2026-07-17. 2026 figures are estimates.
- • 2026-02: combined 2026 guidance crosses ~$700B, nearly double 2025 (CNBC)
- • 2026-06: capex-to-revenue gap widening; markets begin repricing debt-funded buildouts (Forbes)
- • 2026-07-16: SOX enters bear-market territory (-20% from late-June record) on Kimi K3 + Iran escalation
Pipeline decisions
The collector decides autonomously and discloses here — classifications, data gaps, threshold changes, alerts.
- 2026-07-18collectRun complete: 200 models, openShare=74.0
- 2026-07-18collectMarkets as of 2026-07-16: correlated-dd 14.2%, divergence -13.9pp, dispersion -19.6pp
- 2026-07-18collectOpenRouter rankings 2026-07-17: open-weight share 74.0% of top-50 traffic
- 2026-07-18collectOpenRouter: 344 models
- 2026-07-18collectArtificial Analysis: 200 models via https://artificialanalysis.ai/api/v2/language/models/free
- 2026-07-18classificationUnknown creator "Nex AGI" (nex-n2-pro) — defaulted to closed-weights; add to classify.ts if wrong
- 2026-07-18collectRun complete: 200 models, openShare=74.0
- 2026-07-18collectOpenRouter rankings 2026-07-17: open-weight share 74.0% of top-50 traffic
- 2026-07-18collectOpenRouter: 344 models
- 2026-07-18collectArtificial Analysis: 200 models via https://artificialanalysis.ai/api/v2/language/models/free
- 2026-07-18classificationUnknown creator "Nex AGI" (nex-n2-pro) — defaulted to closed-weights; add to classify.ts if wrong
- 2026-07-17configopen-majority trippedAt backdated to 2026-03-19 — the sustained 50% crossing in the measured OpenRouter rankings backfill (first evaluation only saw today)