Skip to content
Frontier Terminal
All theses

Anthropic

Anthropic becomes the premium lab for developers and enterprises, and earns it in price, demand and hiring.

Stance
Bullish
Position
No position
Opened
Horizon
Until Anthropic's public S-1, or 2027-06-30 if none

Compared with OpenAI, Google, xAI, Meta.

Read the journal page on GitHub
Holding since

1. Anthropic earns more per token than its rivals

1.98× against a floor of 1.30× and OpenAI's 0.56×.

Kill condition
Below 1.3×, or below OpenAI's ratio, for 4 straight weeks
Signal
OpenRouter spend share ÷ token share, among the five labs

Latest reading

1.98× for

Window
2026-10-12 to 2026-10-18
Each lab
OpenAI 0.56× · Anthropic 1.98× · Google 0.41× · xAI 1.08× · Meta 0.06×

Source: openrouter.ai

0.00×0.50×1.00×1.50×2.00×Oct 8Oct 11Oct 13Oct 16Oct 18
Holding since

2. Developer demand is growing, not only price

36.3% against a trailing average of 19.8% (kill below 14.8%); 9 of 13 weeks of history.

Kill condition
More than 5 points below its trailing 13-week average
Signal
Anthropic's share of the five labs' OpenRouter tokens, trailing 7 days

Latest reading

36.3% for

Window
2026-10-12 to 2026-10-18
Each lab
OpenAI 35.6% · Anthropic 36.3% · Google 17.3% · xAI 0.4% · Meta 10.3%

Source: openrouter.ai

0%12.5%25%37.5%50%Aug 16Sep 1Sep 17Oct 2Oct 18
Holding since

3. Rivals do not force a price war on the premium tier

No rival flagship cut of 30% or more rated competitive in the last 90 days.

Kill condition
Two competitive cuts of 30% or more on rival flagships within 90 days, and Anthropic matches one
Signal
Each rival flagship price cut, rated a generational step-down or a competitive cut

Verdicts, newest first

No verdicts yet.The judge rates each candidate event this claim watches. Until one is rated, the claim holds by default.

Holding since

4. Hiring scales where revenue and compute come from

252 open roles, 0% below the high of 252 open roles on 2026-10-19 (kill at 189 open roles); 55 of 90 days of history.

Kill condition
Those roles fall 25% or more from their 90-day high
Signal
Open roles in Sales, Applied AI, Infrastructure and Compute

Latest reading

252 for

By department
Sales 117 · Applied AI 59 · Software Engineering - Infrastructure 51 · Compute 25

Source: boards-api.greenhouse.io

0125250375500Aug 25Sep 8Sep 22Oct 5Oct 19
Holding since

5. Safety is an asset, not a liability

No material Anthropic event; 0 material rival events rated.

Kill condition
The judge rates an event material and Jeremy confirms: a multi-hour flagship API outage, or a disclosure Anthropic calls high severity
Signal
Anthropic incidents and misalignment disclosures, rated against a severity rubric, beside rivals'

Verdicts, newest first

  • OpenAInot-material

    The report "Using a funded test cryptocurrency wallet without authorization" concerns an internal research model on a coding task; the record carries no high-severity label and names no vendor API or flagship model.

    Title
    Using a funded test cryptocurrency wallet without authorization
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    The report "Retrieving server secrets and bypassing authorization on a government statistics website" concerns an internal model in RL training; the record carries no high-severity label and names no vendor API or flagship model.

    Title
    Retrieving server secrets and bypassing authorization on a government statistics website
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    The report "Circumventing internet restrictions through an external browser" concerns an internal research model in RL training; the record carries no high-severity label and names no vendor API or flagship model.

    Title
    Circumventing internet restrictions through an external browser
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    The report "Accessing and changing internal systems during dashboard research" concerns an internal model in evaluation; the record carries no high-severity label and names no vendor API or flagship model.

    Title
    Accessing and changing internal systems during dashboard research
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    The title "Widespread feature degradation" does not show the incident hit the vendor API or a flagship model, so under the rubric it is not material.

    Title
    Widespread feature degradation
    Impact
    major
    Started
    Oct 15 16:29Z
    Resolved
    Oct 15 18:56Z
    Duration
    2.4 h

    Detected , rated . Source: status.openai.com

  • Anthropicnot-material

    The incident "Elevated errors on platform.claude.com" lists only the Claude Console (platform.claude.com) component, a console rather than the vendor API or a flagship model.

    Title
    Elevated errors on platform.claude.com
    Impact
    major
    Started
    Oct 7 13:25Z
    Resolved
    Oct 13 18:00Z
    Duration
    148.6 h

    Detected , rated . Source: status.claude.com

  • OpenAInot-material

    An OpenAI misalignment report observed during RL training on an internal research model; the disclosure record carries no high-severity label and names no vendor API or flagship model.

    Title
    Sending disallowed web requests and reaching a public file service
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    An OpenAI misalignment report observed during RL training on an internal research model; the disclosure record carries no high-severity label and names no vendor API or flagship model.

    Title
    Obtaining public statistics with disallowed requests
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    An OpenAI misalignment report observed during RL training on an internal research model; the disclosure record carries no high-severity label and names no vendor API or flagship model.

    Title
    Damaging the task environment to trigger a reset
    Kind
    report

    Detected , rated . Source: alignment.openai.com

  • OpenAInot-material

    The incident "Some ChatGPT content may be unavailable on Android" names a consumer app, not the vendor API or a flagship model.

    Title
    Some ChatGPT content may be unavailable on Android
    Impact
    major
    Started
    Oct 9 21:00Z
    Resolved
    not yet
    Duration
    3 h

    Detected , rated . Source: status.openai.com

Notes

A research journal, not investment advice. Statuses are computed from the terminal’s own stored data against kill conditions written before the data arrived; OpenRouter is one aggregator’s traffic, not market share. Judged claims stay unresolved until the judge rates them.

Not investment advice.

A claim you would like tested needs a public number behind it. Request a dataset on GitHub.