The Innovare Index

Weekly posted-price index for frontier AI capability. Ten flagship models, geometric-mean blended pricing, published weekly. Curated independently by Innovare Melbourne.
15 November 2026 · Edition 023
← innovare.melbourne
⏹ Weekly AI Vital Signs is our weekly companion to this index, tracking the money behind the boom. This week it reads the bill arriving: Nvidia’s founder gone and its first results without him due under an interim chief, Anthropic’s one-point-four-trillion-dollar IPO putting a public price on the whole trade, SoftBank’s seven-billion-dollar data-centre loss, and the quiet tightening of the financing that pays for the build. Open the tracker →
Innovare Index · 15 November 2026
$2.88/ M tokens
▲ Up 27¢ — a premium flagship rolls in
Geometric mean blended USD per million tokens, across 10 flagship models, on the TPI-comparable methodology. The index rose twenty-seven cents this edition, and like the falls before it, the move is composition, not a market verdict: Google’s new premium flagship, Gemini 4 Argon, cleared its four-week general-availability window and rolled into the basket in place of Gemini 3.6 Flash — $8.80 against $3.30, which lifts the geometric mean almost on its own. The irony runs the other way this week — the posted price of frontier AI ticked up, in the same fortnight the market lost its founder-chief at Nvidia and spent its days arguing that the whole thing costs too much to build.
The week the bill came due
The most valuable company on earth is being run by its chief financial officer. Jensen Huang resigned from Nvidia on the twelfth, and by the thirteenth the scandal that took him down had hardened, on New York Times, CNBC and Al Jazeera reporting, into an alleged Chinese-espionage story. In the same week Anthropic put a one-point-four-trillion-dollar price on itself and asked the public to vote, and SoftBank booked a seven-billion-dollar loss on data centres it has not finished building. The question stopped being whether AI works. It became who pays for it.
Basket: Gemini 4 Argon ($4/$20, $8.80 blended) rolls in for Gemini 3.6 Flash after clearing the GA window. GPT-6.1 Terra, Kimi K3.1 Pro and GPT-6 Astra stay under review. ⏹
This week
Edition 023 is the week the bill for the AI boom stopped being a worry and became a line item. The most valuable company on earth lost its founder — and by midweek the scandal that took him down had hardened, on major-outlet reporting, into an alleged espionage story. Anthropic put a one-point-four-trillion-dollar price on itself and dared the public to pay it. SoftBank booked a seven-billion-dollar loss on data centres it hasn’t finished building. A coalition took five of the biggest labs to court to try to make them stop. And underneath all of it, the money that funds the build quietly got harder to raise. Last edition the question was who pays. This week, several different people started sending the invoice.

MarketsNvidia after Jensen — and the scandal becomes a spy story — 11–13 November 2026

On the twelfth of November, Jensen Huang resigned as chief executive of Nvidia, and Colette Kress, the company’s chief financial officer, took over as interim head of the most valuable company on earth. For a week the story had been an affair and a boardroom mess. Then, on the thirteenth, the New York Times, CNBC and Al Jazeera put reporting behind it, and it became something far heavier. U.S. agencies had concluded that the woman at the centre of it — Alice Huang, a Megaspeed executive, no relation to Jensen — was aiding Chinese intelligence, and that her firm had been diverting banned Nvidia chips into China in violation of export controls.

Read it slowly, because the fairness matters. Intelligence also concluded that Jensen Huang himself did not know her role — a point we carry deliberately. What makes this a Washington story rather than a tabloid one is the timeline the Times reports: a secret memo in the spring of 2026, its drafting reportedly supervised by Secretary of State Marco Rubio, warned she was a spy for China’s Ministry of State Security; the President was briefed in late 2025 and chose not to act; and the federal case moved from Commerce to the Justice Department this month, after an earlier Commerce probe was reportedly halted when Nvidia’s own lawyer intervened. Senator Ron Wyden and Representative Richard Neal have opened an investigation into how the administration handled it, and Huang has agreed to testify to Congress before Thanksgiving.

The honest caveat, because we insist on it: none of the espionage allegations have been adjudicated, the figures around Megaspeed’s chip purchases are secondary and contested, she denies wrongdoing, and the Justice Department says it has reached no conclusions. But strip the scandal away and the structural fact is the one this index keeps returning to: the entire AI build rests on a single company’s chips, that company has just lost the man who was its whole strategy, and its first quarterly results without him land this week under a chief executive doing the job for the first time.

FinanceAnthropic puts a $1.4 trillion price on itself — 10 November 2026

Two weeks after filing, Anthropic set the terms: seventy to seventy-eight dollars a share, a fully diluted valuation of roughly one-point-four trillion dollars, and as much as sixty-four billion dollars raised — a roadshow that began the week of the ninth and a debut expected before Thanksgiving. The number is staggering and, in context, a climbdown: only weeks ago bankers were reportedly floating two trillion. The market is still willing to buy the AI story, in other words — just not at the price it was dreaming about in October.

This is the cleanest test we have of what public investors actually think a model-maker is worth, and it lands in the same fortnight the stock market convulsed over AI spending and Nvidia lost its chief. Reporting since suggests the order book is heavily oversubscribed, which cuts against the simple bubble-popping read: a 1.4-trillion-dollar company losing money today is a bet that the revenue curve eventually climbs to meet the spending. Disclosure, as always: we are an Anthropic customer and this index is written with their model, so weigh us accordingly.

FinanceSoftBank’s $7 billion loss — the build-out shows up as a number — 10 November 2026

SoftBank reported a seven-billion-dollar quarterly net loss, and the cause is almost poetically on-theme: a three-point-seven-billion-dollar writedown on data-centre power equipment, taken after its US plans changed and left it holding gear it had already paid for in advance. Financing costs more than doubled over the quarter. The company noted that some future purchases could still qualify for up to two-point-one billion dollars in rebates — a silver lining on a very expensive cloud.

This is the most tangible version of the week’s whole argument. The AI build-out is usually described in commitments and forecasts; here it turned into a realised loss on real equipment, bought ahead of plans that then shifted. It is the cost of moving first in a market where the ground keeps moving — and a reminder that “capex” is not an abstraction. It is machines someone has already paid for, whether or not the building meant to hold them ever gets its permit.

EconomicsThe money gets harder — financing the build hits friction — 11–14 November 2026

Underneath the headlines, the plumbing tightened. Bloomberg reported that Blackstone’s QTS, one of the largest data-centre operators, chose to raise a smaller private-placement deal rather than borrow a much larger sum at today’s higher cost — a refinancing that resizes rather than defaults, but a telling choice all the same. Separately, the Banker reported that lenders are increasingly using credit insurance to offload part of their data-centre exposure and free up room to keep lending.

Neither is a crisis, and we won’t dress them as one — the QTS figure is single-sourced, and a smaller deal is not a failed one. But together they sketch, from the financing side, the same picture SoftBank sketched from the spending side: the cheap, limitless money that underwrote the first phase of the build is getting choosier. And the honest worry with the insurance trick is that it does not remove the risk, it moves it — from the banks that understand these loans to the insurers who may not.

RegulationA coalition sues five AI labs to make them stop — 12 November 2026

On the twelfth, a coalition of non-profits filed suit against Anthropic, Google, Meta, OpenAI and SpaceX’s AI division, asking a court to halt the development of more advanced systems on the grounds that they pose catastrophic risks — the plaintiffs point specifically to bioweapons and cyberattacks. It is, to be clear, a request for a court to intervene, not an order that stops anything; the odds of a judge freezing the US AI industry by injunction are long.

What earns it the slot is not the likelihood of winning — it’s that the catastrophic-risk argument has graduated from open letters and conference panels into a filed legal complaint naming the five biggest builders at once. Whatever the court does with it, that is a new front. The near-term business consequence is uncertain and probably small; the signal — that “slow down” now has a docket number — is not.

Editorial method & sources for this edition. Analysis and framing are editorial synthesis by Innovare Melbourne. We hold an optimistic-skeptic view: this technology is genuinely useful, genuinely disruptive, and will genuinely hurt people — all three at once. Every claim is checked against at least two independent sources before it runs, and figures that differ across sources are given as ranges, not silently reconciled. On the Nvidia story we draw a hard line: the resignation and the interim-CEO transition are on the public record (the 8-K), but every element of the alleged Chinese-espionage case is attributed and unadjudicated — intelligence reportedly concluded Jensen Huang himself did not know Alice Huang’s role, she denies wrongdoing, and the Justice Department says it has reached no conclusions; the Megaspeed chip figures are secondary and contested. Anthropic’s “$1.4 trillion” is the fully-diluted figure, not reconciled to the $70–78 price range on the filing’s post-offering share count. SoftBank’s ~$7 billion loss and $3.7 billion power-equipment writedown are taken from its own Q2 report. The QTS financing figure is single-sourced and flagged as such. Gemini 4 Argon is counted at its $4/$20 permanent list, not its $2/$10 introductory rate. Disclosure: Innovare is an Anthropic customer and this index is produced with an Anthropic model; that interest touches both the Anthropic IPO item and the coalition-lawsuit item above. Corrections: aaron@innovare.technology.

News sources cited this week:
· Nvidia — Huang resignation accepted, Colette Kress interim CEO — NVDA 8-K; CEO-transition exhibit; Nvidia board statement (11 Nov)
· Nvidia — the Alice Huang / Megaspeed espionage reporting (attributed, unadjudicated) — CNBC; Axios; Al Jazeera; NYT via TipRanks; Fortune (9–12 Nov)
· Congressional investigation — Wyden & Neal on the chip-sale-cut arrangement — Senate Finance (Wyden) (Nov 2026)
· Market reaction — AI stocks sink on the exit — Kitco/Reuters (12 Nov)
· Anthropic IPO — $70–78/share, ~$1.4T fully diluted, debut before Thanksgiving — S-1/A no. 2 (SEC); Decrypt/CNBC (10 Nov)
· SoftBank — ~$7B quarterly net loss, $3.7B data-centre power-equipment writedown — SoftBank Q2 FY26 report; earnings-call transcript (10 Nov)
· Financing strain — QTS smaller private placement (single-sourced, flagged); lenders using credit insurance — Bloomberg; the Banker (11–14 Nov)
· Coalition lawsuit — four nonprofits sue five AI labs to halt frontier development — KuCoin; Crypto Briefing; Tech Times UK (12 Nov)
· Index pricing — Gemini 4 Argon $4/$20 standing list ($8.80 blended), $2/$10 intro through 30 Nov — Mixed-News; FourWeekMBA; provider pricing pages (Anthropic, OpenAI, Google, Moonshot, Alibaba Model Studio) cross-checked against the public Token Price Index, verified 15 November 2026
Index members · 15 November 2026
▶ The index signal

$2.88, up twenty-seven cents — and once again the move is composition, not a market verdict. One thing changed. Google’s new premium flagship, Gemini 4 Argon, cleared its four-week general-availability window and rolled into the basket in place of Gemini 3.6 Flash — at $8.80 against $3.30, a premium model displacing a mid-tier one, which lifts the geometric mean almost on its own. Every other posted list price held. The irony runs the other way this fortnight: the posted price of frontier AI ticked up, in the same days the market lost Nvidia’s founder and spent itself arguing the whole thing costs too much to build.

The structural fact underneath still holds. The cheap tier — two models at twenty-two cents — is where the floor keeps dropping, and Chinese open-weight models still carry more than half of global token consumption. The index rose this week only because a dearer flagship entered, not because anyone’s bill went up. Cheaper to run every week; the entire argument is still about what it costs to build.

One move this edition. Gemini 4 Argon ($4/$20, $8.80 blended) rolls into the basket in place of Gemini 3.6 Flash ($3.30 blended) after clearing the four-week GA window. Every other member held its posted list price. The index rises to $2.88, up about twenty-seven cents — composition, not a market-wide repricing. Verified against provider pricing pages and cross-checked with the Token Price Index, 15 November 2026.
Composition: Argon actioned, the rest of the wave still held. The next successors stay under the four-week window, with prices already set: GPT-6.1 Terra ($1/$5, $2.20 blended) and Kimi K3.1 Pro ($2/$10, $4.40 blended) clear in late November. GPT-6 Astra ($10/$50, $22.00 blended) remains eligible but is not actioned. We report each roll as its window clears, rather than churning the basket on launch-week noise.
Two lines we could not confirm to a provider page this week and are therefore holding at last edition’s value, flagged for correction: GPT-5.6 Terra remains off OpenAI’s public pricing page as its GPT-6.1 Terra successor (in-window) takes over, so its $5.00 is carried from the Token Price Index; and Moonshot’s Kimi K2.6 page still renders empty, with trackers split between $1.87 and $1.57 — we keep $1.87 pending a provider confirm. New entrant Gemini 4 Argon bills a $2/$10 introductory rate through 30 November 2026; per methodology the basket uses its $4/$20 standing list ($8.80 blended), with the intro noted.
ModelProviderBlended $/MLN(blended)vs Ed.022
FrontierClaude Fable 5.1Anthropic$22.003.091▶ flat
PremiumClaude Opus 5.5Anthropic$8.802.175▶ flat
MidGPT-5.6 TerraOpenAI$5.001.609▶ held
MidGPT-6.1 SolOpenAI$4.401.482▶ flat
MidClaude Sonnet 5 ❅Anthropic$4.401.482❅ frozen
PremiumGemini 4 ArgonGoogle$8.802.175▲ new (in for Flash)
CheapQwen3.7 MaxAlibaba$2.640.971▶ flat
CheapClaude Haiku 5.5Anthropic$0.22-1.514▶ flat
CheapKimi K2.6 (open)Moonshot AI$1.870.626▶ held
CheapGPT-6 LunaOpenAI$0.22-1.514▶ flat
❅ Claude Sonnet 5 pricing ($2/$10, $4.40 blended) made permanent on 10 August 2026. Blended = input × 0.7 + output × 0.3. Cache hits excluded for reproducibility. This edition’s move: Gemini 4 Argon ($4/$20, $8.80 blended) replaces Gemini 3.6 Flash ($3.30 blended) after clearing the four-week GA window (Argon went GA 15 October). Argon’s $2/$10 introductory rate (through 30 November 2026) is noted but not counted; the basket uses its $4/$20 standing list, per methodology. Qwen3.7 Max: international list $1.65/$4.951, blended $2.64. Two lines held and flagged this week: GPT-5.6 Terra ($5.00, delisted from OpenAI’s page, carried from TPI) and Kimi K2.6 ($1.87, provider table empty, trackers split $1.87/$1.57). All other rates verified against provider pricing pages and cross-checked with the Token Price Index, 15 November 2026.
Basket review

Argon lands — a premium flagship enters the basket

Gemini 4 Argon went to general availability on 15 October and has now cleared the standing four-week window, so it rolls into the basket in place of Gemini 3.6 Flash — a premium flagship displacing a mid-tier model, and the whole of this edition’s twenty-seven-cent rise. The rest of the recent wave stays under the same test, with prices ready for the roll the moment each clears. We report every decision as it lands, rather than churning the basket on launch-week noise.

ModelProviderInput $/MOutput $/MBlended $/MStatus
Gemini 4 ArgonGoogle$4.00$20.00$8.80✓ Actioned — rolls in for Gemini 3.6 Flash
GPT-6.1 TerraOpenAI$1.00$5.00$2.20Under review (clears ~late-Nov)
Kimi K3.1 ProMoonshot AI$2.00$10.00$4.40Under review (clears ~late-Nov)
GPT-6 AstraOpenAI$10.00$50.00$22.00Under review (full GA)
The Argon roll, all else equal, lifts the geometric mean by about twenty-seven cents — composition, not a market-wide move. The remaining composition question — whether GPT-6 Astra should enter and which older member it would displace — is held to the GA window and will be shown with the arithmetic when it lands. Editorial note: Argon-for-Flash is a provider-for-provider roll (Google’s newest flagship in for its older basket member); the displacement is the one judgment in this edition’s index.
Updated this week

⏹ AI Vital Signs — the financials, tracked

This fortnight’s money story — the founder of the company the whole build rests on gone and about $1.4 trillion off Nvidia since its October record, four central banks warning about the same thing in a single week, and one IPO filing putting a public price on the trade — is exactly what the price table alone cannot show. AI Vital Signs is our standing, weekly-updated dashboard of the money behind AI: the total capital at risk, the tower of debt and off-balance-sheet commitments, US versus China, the labs’ revenue against what they have promised to spend, and the circular financing that ties it all together. New on the board this fortnight: the Federal Reserve’s Vice Chair for Supervision putting AI-related debt “on track to exceed $800 billion this year” with off-balance-sheet commitments estimated at “more than three times” that, and naming “circular financing” outright — echoed by the ECB, the Bank of England and the RBA in the same week; Anthropic’s $70–78 IPO range (~$1.4 trillion fully diluted), with $26 billion of stock bought by its own chip and cloud suppliers; SoftBank’s leverage now partly disclosed — group loan-to-value 15.9%, net asset value down about a quarter in six weeks, a $3.7 billion writedown on data-centre power gear; and OpenAI’s first close of about $11 billion at a $1.25 trillion valuation, roughly a third of its original target.

Nvidia −$1.4T Circular financing — named Anthropic $1.4T SoftBank levered Revenue vs spend The debt tower
Open AI Vital Signs →
Layoff & Re-hire Tracker · 2026 · Updated Edition 023

Jobs lost and roles reversed

A curated log of significant workforce reductions where AI was cited as a factor, and documented reversals where roles cut for AI reasons were subsequently reinstated. Updated each edition.

Losses · recent weeks to 15 November 2026

CompanyHeadcountAI citedDateContext
Amazon ~14,000 ✓ Yes Oct 2026 Corporate roles cut — the company's reporting tied it directly to AI “taking over routine work.” The largest single AI-cited corporate cut of the year, and Amazon alone now accounts for roughly half of 2026's tech layoffs. Reuters; NewsNation.
Canva (AU) ~1,271 ~ Partial 4 Nov 2026 The Melbourne company's first-ever major restructure, framed by its CEO as a reset after its “hardest year yet.” The honest read: the Financial Review reports the cuts largely target AI-related roles rather than AI replacing workers — a reshuffle around AI, not a clean replacement story. Forbes Australia; HCA Mag; ACS.
Barclays (UK) ~5,000 (~5%) ~ Partial 22 Oct 2026 A bank cutting about five per cent of its staff in a cost drive. The honest read is in the CEO's own framing: he called AI the “icing on the cake” — the restructuring is the cake, AI the accelerant. The clearest sign yet that banks are starting to bank the efficiency. ITV; Reuters via Business Day; FinTech Futures.

Re-hires & reversals

CompanyRoles cutStatusDateWhat happened
Commonwealth Bank (AU) ~40–45 ✓ Rehired 2026 AI voice bot deployed as replacement generated a surge in call volume rather than reducing it; roles reinstated; CBA issued a formal acknowledgement. The documented case that started this tracker.
Running totals (Challenger, Gray & Christmas October report, released 5 November — still the latest; the November report is due early December): AI has been cited in 121,401 US job-cut announcements so far in 2026, about twenty per cent of all cuts — but in October alone AI was cited for just 1,265 cuts, its lowest monthly total this year, and “market and economic conditions” has now overtaken AI as the leading stated reason for 2026. US employers announced 617,189 cuts year-to-date. The monthly cooling while the year-to-date total keeps climbing is worth holding both ways. No new AI-cited layoff cleared our two-source bar in the fortnight to 15 November, so the rows above stand. As before, announced cuts where employers cited AI are not proof of completed displacement, and jobs never posted are not counted here.
AI Reality

Watch the series

Short-form video essays on AI in business — the thinking behind this index. Evidence, not hype.

Methodology

Innovare Index = exp((1/n) × Σ ln(blended_price_i)),  n = 10

The Index is the geometric mean of the ten member models' blended prices, expressed in USD per million tokens. The geometric mean treats proportional differences equally — a model twice as expensive as another carries the same weight whether the comparison is $0.20 vs $0.40 or $20 vs $40. This produces a more balanced picture than an arithmetic mean, which is dominated by the most expensive entries.

The blended price for each model is input × 0.7 + output × 0.3. This is the same formula as the public Token Price Index. Cache hit pricing is excluded because cache-hit rate is a workload assumption — including it would mean the Index measures workload patterns as much as posted prices. Temporary promotional pricing is also excluded: the basket uses standing list prices, so a model's entry reflects a durable posted price rather than a limited-time offer. The Index is a posted-price market reference, not a forecast of any organisation's actual bill.

The Index rises when the market composition shifts toward more expensive models, or when individual member prices rise. It falls when cheaper models enter or when prices fall. Mix changes are intentional — the Index tracks the frontier as the market defines it, not a fixed historical basket. A flat edition means every posted list price held and the basket was unchanged; this edition was not flat, because Google’s Gemini 4 Argon ($8.80 blended) rolled into the basket in place of Gemini 3.6 Flash ($3.30) after clearing the four-week general-availability window, lifting the Index about twenty-seven cents even as every other list price held.

What the Index does not measure

Posted prices are not realised costs. Organisations typically pay less because of enterprise discounts, prompt caching, model routing, batch processing, and committed infrastructure spend. The Index is the published list price; a separate practitioner calculation incorporating cache hits and other discounts can run 50–70% lower for production workloads with prompt caching configured.

Inclusion criteria

The Innovare Index is curated to ten models drawn from at least five providers, with at least one entry in each capability tier (Frontier, Premium, Mid, Cheap). To enter the Index, a model must meet all six criteria:

01
Commercial API availability
Accessible via a paid, publicly available API with stable USD-per-million-token pricing. Free preview tiers excluded until commercial pricing publishes.
02
General-purpose capability
Supports general text generation across at least three of: summarisation, generation, reasoning, coding, retrieval. Specialist-only models excluded.
03
Market presence
Documented adoption through enterprise availability, presence in independent benchmarks, or flagship recognition from its provider.
04
Not scheduled for deprecation
Active and supported. When a provider releases a successor at similar price, the Index entry updates at the next weekly review, subject to a four-week general-availability window.
05
Editorial relevance
Models that buyers in mid-market production environments are actively choosing between. This criterion is more opinionated than the public TPI; the Index is curated, not comprehensive.
06
Provider diversity
Members drawn from a minimum of five providers. Anthropic is currently over-weighted (40%) because the editor's stack runs Anthropic at the production tier — this is disclosed editorial bias.

Data sources

Token pricing
Pulled each Saturday from provider pricing pages: Anthropic, OpenAI, Google AI for Developers, Moonshot, Alibaba Cloud.
Intelligence Index & Rank
Artificial Analysis — independent third-party benchmarking. Index v4.0 aggregates 10 evaluations: GDPval-AA, τ²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt.
Geometric-mean methodology
Adapted from the public Token Price Index (tokenpriceindex.com), which tracks 19 models. The Innovare Index uses the same blended-price formula so values are directly comparable, with editorial divergence on model selection.
Historical snapshots
Each edition's rates archived to an internal Notion database; full reproducibility for any past Index value on request.

About Innovare

IM
Innovare Melbourne
Building AI-native software for mid-market clients

Innovare builds AI-native systems for mid-market organisations. Practitioner-led, augmentation-not-replacement stance, opinionated about where current models earn their keep and where they don't. The Innovare Index is the public artefact of a weekly internal pricing review and was inspired by the methodology pioneered at tokenpriceindex.com.

Disclaimer

The Innovare Index is a posted-price reference for AI inference, compiled from publicly available provider pricing. It is not financial, investment, or procurement advice. Realised costs depend on workload, caching, batch processing, enterprise agreements, and other factors not modelled here. The Index is a curated subset; no model selection methodology is perfectly objective. Methodology and inclusion criteria are documented above so any value can be reproduced independently. Company holdings and legal filings referenced in this edition are reported from public disclosures and as allegations where not yet adjudicated, not as findings or as financial advice. Errors and corrections welcome — contact via Innovare Melbourne.