All sources

What it says: $65 billion Series H at a $965 billion post-money valuation, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital.

Cited on:Claude

What it says: Anthropic is a public benefit corporation; its Board of Directors is elected by stockholders and by its Long-Term Benefit Trust. The page does not state founding year, founders, or headquarters.

Cited on:Claude

What it says: Claude Code is an agentic coding tool that reads your codebase, edits files, runs commands, and integrates with development tools, on terminal, IDE, desktop app, and web. Most surfaces require a Claude subscription or Anthropic Console account; it handles git commits and pull requests and connects to tools through MCP. Undated page, checked today.

Cited on:Claude

What it says: Lists Ultra at $200 a month: everything in Max plus no 5-hour limit for verified accounts, 10x Pro usage every week, and $200 a month in API credits (up from $100 on Max). Max is $100 a month with 5x Pro usage, Fable up to half the weekly limit, early access, and priority access; Pro includes Claude Code, Research, and Projects. The page does not mention a 20x tier.

Cited on:Claude, Home

What it says: Cloudflare agreed to acquire Replicate; Ben Firshman is named co-founder and CEO; financial terms not disclosed; expected to close within two months.

Cited on:Replicate

What it says: News list carries a September 30, 2026 headline that ElevenLabs' valuation increased to $22 billion. Founders, founding year, and headquarters are not stated on the page. The page title was not captured by the fetch; the title shown is descriptive.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: Eleven v4 is "Our most emotive, high quality speech synthesis model" (90+ languages); Flash v2.5 targets real-time use at about 100 ms median latency; Multilingual v2 is the most stable for long-form. Character limits run 5,000 to 40,000; one credit per input character; unused credits roll over up to two months; voice library of more than 10,000 voices. Undated, checked today.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: Scale is $299 a month: 1.8M credits a month, 3 workspace seats, everything in Pro plus Team Collaboration and 3 Professional Voice Clones. Pro is $99 a month (600k credits); Business is $990 a month (6M credits, 10 seats).

Cited on:Voice: ElevenLabs, Rime, and Kokoro, Home

What it says: Exa describes itself as a search index for AI agents with Search, Contents, Agent, and Monitors APIs, Websets, and an MCP server; "Join 900,000+ developers." Claims Exa 3.0 scores 91.8% on BrowseComp versus Perplexity 91.5%, Brave 83.5%, and Parallel 72.7% (self-reported). Security: SOC 2 Type II, GDPR, CCPA, HIPAA with a BAA, and zero data retention. Gives no latency figure on the page.

Cited on:Exa

What it says: Lists Will Bryk as CEO and Jeff Wang as co-founder; footer reads EXA LABS INC. Founding year and headquarters are not stated.

Cited on:Exa

What it says: Free Tier $0 with $10 in credits (up to 2,500 Lite searches), resetting to $10 on the first of each month. Pay-as-you-go: "no subscription and no minimum spend." Per 1,000 requests: Search lite/instant $4, auto/fast $7, Deep Search deep-lite/deep $12, deep-reasoning $15; Contents $1 per 1,000 pages; Answer $5; Monitors $2; results above 10 cost $1 per 1,000. Agent API fixed-effort prices run $0.012 to $1.00 per request. Enterprise is custom; a dashboard option costs $12,000 upfront and includes $12,000 in credits and 25 QPS.

Cited on:Exa, Home

What it says: "Exa just passed $100M in annualized revenue" and now handles "2 billion requests a month from 200k+ teams" (company claims). No headquarters or headcount stated.

Cited on:Exa

What it says: $250 million Series C at a $2.2 billion valuation, led by a16z. Says Exa "already powers search for Cursor, Cognition, HubSpot, OpenRouter, Monday.com and over 400,000 developers," and "We started Exa five years ago." Will Bryk is CEO and co-founder. No headquarters stated.

Cited on:Exa

What it says: $85 million Series B at a $700 million valuation, led by Benchmark, with Lightspeed, Y Combinator, and NVentures (NVIDIA's venture arm).

Cited on:Exa

What it says: Exa Search is a web search API returning ranked results with page content (highlights, text, or summary). Types: auto ~1 second, lite ~450 ms, deep-lite ~4 seconds, deep 4-15 seconds (approximate, cached content). Up to 10 results by default, up to 100 via numResults, no pagination; lite never fetches a page live and rejects outputSchema and several other parameters; summaries cost extra; no rate limits stated. Undated; the page title was not captured, so the title shown is descriptive.

Cited on:Exa

What it says: "Google Antigravity is our agentic development platform." Products: App, CLI ("Run autonomous coding agents, execute shell commands directly"), IDE ("The fully-featured, agentic IDE"), SDK, and Extensions. The page documents no product limits.

Cited on:Google AI Ultra, Gemini, and Antigravity

What it says: Individuals tier is $0 a month with basic weekly rate limits. Google AI Ultra adds higher access to the latest Gemini models, more rate limits, and a flexible AI credit pool; no Ultra price is listed on this page.

Cited on:Google AI Ultra, Gemini, and Antigravity

What it says: Google AI Ultra: "Starting at: $99.99 / month" and "$199.99 / month" for the higher tier. Includes the highest access to Gemini 4 Argon, Google Flow credits, higher Antigravity agent rate limits (5x or 20x by tier), YouTube Premium individual, Google Home Premium Advanced, Project Genie, and storage starting at 20 TB. Gemini app usage described as up to 20x higher than Pro.

Cited on:Google AI Ultra, Gemini, and Antigravity, Home

What it says: Names the tiers Ultra 5x and Ultra 20x: 20 TB vs 30 TB storage and 10,000 vs 25,000 Flow credits a month. Says Ultra offers "5x higher usage limits in Gemini than the AI Pro plan." The Ultra price figure was missing from the fetched text; the savings calculator listed Ultra at $99.99 a month.

Cited on:Google AI Ultra, Gemini, and Antigravity

What it says: An open-weight TTS model with 82 million parameters, Apache-licensed. Architecture from Li et al. (StyleTTS 2); trained by @rzvzn on Discord. v1.0 published 2025-01-27 with 8 languages and 54 voices; v0.19 (2024-12-25) had 10 voices.

Cited on:Voice: ElevenLabs, Rime, and Kokoro, Home

What it says: Lists two paid Muse plans: Power at $20/month with "500M Muse tokens per week" and Maximum at $100/month with "3B Muse tokens per week"; Muse is also free with a usage limit (no figure given). The page never uses the name "Max" for a Muse plan and shows no currency code or token definition. Renews automatically each month; sold at muse.ai and in the mobile app; "Muse and Muse subscriptions are in limited testing and aren't available in all locations yet." Undated, checked today.

Cited on:Muse (Meta), Home

What it says: Muse is a personal AI agent from Meta: you give it a goal or task and it works on it (email, reservations, shopping, budgeting, calendars, fitness plans). Each user gets a cloud computer: "Muse Secure VM is a persistent, isolated Linux computer with a full browser." The page lists no plans or prices.

Cited on:Muse (Meta)

What it says: Published September 15, 2026 (updated September 16). Meta One plans: Core $7.99/mo, Premium $19.99/mo, Essential from $14.99, Advanced from $49.99, Expert from $149, and Max from $499 a month (Expert and Max are creator and business bundles: "our highest levels of feature access and Meta Business Agent capacity for teams managing at scale"). The page does not label the currency and says plans "may vary by region." The title shown comes from the search listing.

Cited on:Muse (Meta)

What it says: Published April 8, 2026 (updated May 12). Muse Spark is "the first in a new series of large language models built by Meta Superintelligence Labs." Meta says: "This initial model is small and fast by design" and "Muse Spark is an early data point on our trajectory." API access was limited to a private preview for select partners. The title shown comes from the search listing; the fetch did not return the page title.

Cited on:Muse (Meta)

What it says: Published October 12, 2026 (also shows October 13). Muse Wonder and Muse Wonder Full are "our most capable models to date," built on Muse Spark, aimed at agentic tasks and expert domains including law, finance, and health. Muse Wonder will "soon power experiences in Muse," rolling out first to subscribers. "A full report with additional safety evaluations will be available soon."

Cited on:Muse (Meta)

What it says: Published September 8, 2026 (updated September 30). "Muse is a personal AI agent. It doesn't just answer questions, it actually does the work." Powered by Muse Spark; runs on Muse Secure VM; checks with the person before sensitive actions such as sending an email or making a purchase; a separate Sentinel agent must approve traffic to the internet; has no visibility into passwords or payment methods. "It's free for most of what people need, with subscription plans for people who want to do more." Rolling out in the US on iOS, Android, and muse.ai.

Cited on:Muse (Meta)

What it says: Codex surfaces: ChatGPT desktop, mobile, and web apps, the Codex CLI, the Codex IDE extension, and Codex Cloud. Starting tasks: explore and understand code, build a new feature, review code, fix issues and failures. Limits are linked to separate pages (permissions, sandboxing, approvals). Reached via redirect from developers.openai.com/codex.

Cited on:ChatGPT Pro and Codex

What it says: Pro is labeled "Best for research and coding" with "Your choice of 3 usage tiers." It lists Pro reasoning powered by GPT-6 Astra, longer Codex and ChatGPT Work sessions, unlimited and faster image creation, maximum deep research, and maximum memory and context. The fetched text omitted every dollar figure.

Cited on:ChatGPT Pro and Codex

What it says: Pro comes in "$100, $200, or $500 USD per month" plans. Codex is included on Free, Go, Plus, Pro, Business, Edu, and Enterprise. Fast mode counts 2.5x and the Ultrafast modes 8x against included usage; Ultrafast is on Pro $500 and eligible Enterprise/Edu plans. Reached via redirect from developers.openai.com/codex/pricing.

Cited on:ChatGPT Pro and Codex, Home

What it says: Replicate lets you "run AI models with a cloud API, without having to understand machine learning or manage your own infrastructure." Examples cover image, video, and audio; Node.js and Python clients, HTTP API, streaming, webhooks. Links to rate limit and data retention pages without numbers. Undated, checked today.

Cited on:Replicate

What it says: Most public models bill by run time at a per-second hardware rate; some bill per unit (flux-1.1-pro $0.04 per image, flux-dev $0.025, wan-2.1-i2v-480p $0.09 per second, wan-2.1-i2v-720p $0.25 per second). T4 $0.81 an hour, H100 $5.49 an hour. Private models bill for all online time including idle.

Cited on:Replicate, Home

What it says: Replicate will keep operating as its own brand; "The API isn't changing." Replicate grew out of Cog, an open-source tool for defining models.

Cited on:Replicate

What it says: Free $0 (no card; $50 of usage a month, no rollover). Flex: no monthly fee, billed by the minute. Scale: $100/month with $100 of credit a month, rolling over up to $200. Enterprise: custom, billed annually. Per minute, Flex / Scale: Mist v3 text to speech $0.030 / $0.024, Coda v2 text to speech $0.050 / $0.040, Prism speech to speech $0.080 / $0.064, Rime STT $0.002 / $0.0016 (free until Nov 19). Text to speech is "metered by characters generated, about 1,000 characters per minute." Coda v2 lists 313 curated voices, Mist v3 78; custom voices are Enterprise-only. No currency label; commercial use not addressed.

Cited on:Voice: ElevenLabs, Rime, and Kokoro, Home

What it says: "Voice models made for human conversation." Models: Prism (speech to speech, "570ms TTFA"), Coda and Mist (text to speech, "Sub-100ms time to first byte", "600+ voices, 50+ languages"), and Rime STT. SOC 2 Type II with a BAA available; VPC or on-prem deployment. Says it powers "over 1.5 million minutes of conversation." Customer logos include Twilio, Dialpad, LiveKit, and Mayo Clinic. Latency and scale figures are Rime's own claims.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: $24 million Series A led by M13, with Corazon Capital, Unusual Ventures, Cadenza Ventures, and Twilio Ventures. Lily Clifford is CEO and Co-Founder; Morgan Blumberg of M13 joins the board; Rafael Valle joins as Chief Scientist. Location not stated.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: "On December 1, 2025, the Company acquired all of the outstanding shares of Replicate" for total purchase cash consideration of $57.4 million. Date shown is the period end; the filing date was not on the page I read.

Cited on:Replicate

What it says: "We were incorporated in Delaware in July 2004." Principal offices at 1 Meta Way, Menlo Park, California 94025. Mission: "build the future of human connection and the technology that makes it possible." Class A stock trades on Nasdaq as META; as of January 23, 2026, 2,187,177,748 Class A and 342,377,716 Class B shares outstanding. "Mark Zuckerberg, our founder, Chairman, and CEO" can exercise a majority of the voting power. Date shown is the period end; the filing date was not on the part of the page I read (about 200,000 of 402,371 characters).

Cited on:Muse (Meta)

What it says: Q3 2026 revenue $64.6 billion, up 26% year over year; net income $9.86 billion including a $9.89 billion charge tied to the agreement with U.S. State Attorneys General; headcount 66,775 at September 30, 2026; 2026 capital expenditures expected at $135-145 billion. Zuckerberg cites the launch of Muse and Muse Wonder, "our first frontier model." The CFO says MSL investments "will take longer to generate returns than those in our core business."

Cited on:Muse (Meta)

What it says: "Suno is a music company building the world's first creative entertainment platform." Founded 2022; works across 4 offices. Names no founders and no headquarters.

Cited on:Suno

What it says: Monthly and Yearly toggle ("Save 20%" on Yearly). Free $0: 50 credits per day, free models, no commercial rights, no monthly song downloads. Pro $10/month month-to-month ($8/month billed $96/year): 2,500 credits and 20 song downloads per month, best models, commercial use rights. Premier $30/month ($24/month billed $288/year): 10,000 credits and 60 song downloads per month, best models, commercial use rights, Suno Studio access. Subscription credits "do not carry over from day to day or month to month." "See the terms of service for limitations on commercial use."

Cited on:Suno, Home

What it says: Partnership with Warner Music Group to build new models using "high-quality licensed music"; WMG artists may opt in for use of their names, images, likenesses, voices, and compositions. Free users can't download; downloading requires a paid account. Does not mention Songkick.

Cited on:Suno

What it says: v6 is a new generation of Suno models developed with partners including Warner Music Group, BMG, and Believe: v6 (flagship), v6-wild ("less predictable and more varied"), v6-mini (faster). v6 and v6-wild are for Pro and Premier subscribers; v6-mini is for everyone. Features: edit one section or one lyric, mashups, sampling a riff or instrument, starting from text, audio, images, or video. Suno will "retire our previous models." No length caps, quotas, or benchmarks stated.

Cited on:Suno

What it says: $250 million Series C at a $2.45 billion post-money valuation, led by Menlo Ventures, with NVentures, Hallwood Media, Lightspeed, and Matrix. Says "nearly 100 million people have made music on Suno." By Mikey Shulman, Co-Founder & CEO.

Cited on:Suno

What it says: Over $400 million Series D at a $5.4 billion post-money valuation, led by Bond Capital; new investors IVP, Forerunner, Union Square Ventures, Alkeon, and Quiet; existing Matrix, Lightspeed, Menlo Ventures, and Schroders Capital. No revenue or user figures.

Cited on:Suno

What it says: $17 million Series A (total $22 million with a $5 million seed) from Lightspeed, Nvidia's NVentures, and Y Combinator (summer 2021 cohort). Founders Will Bryk (CEO) and Jeff Wang (co-founder), who met as Harvard freshmen. Says Exa "is located in San Francisco."

Cited on:Exa

What it says: Rime was founded in 2022 by Lily Clifford (former Stanford PhD student), Brooke Larson (former Amazon Alexa engineer), and Ares Geovanos (Stanford engineer); $24 million Series A led by M13 Ventures; earlier $5.5 million seed; located in San Francisco.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: Jev returns typed probabilistic outputs; claimed 70ms-500ms response time; pricing $0.042/MTok input and $0 output. TypeSafe has $40 million in funding; Diogo Almeida, a former OpenAI researcher, is co-founder and CEO. Notes TypeSafe's hallucination-free claim is an unfair comparison. Does not state headquarters.

Cited on:Jev (TypeSafe)

What it says: TypeSafe is "an AI lab building machine-native intelligence infrastructure for automation"; Jev is in early access. Lists "$42" per billion input tokens. No plans or tiers listed; founders and headquarters not stated.

Cited on:Jev (TypeSafe), Home

What it says: Jev is TypeSafe's flagship model and the first System One model: you send a state plus typed questions and get structured typed answers with probabilities, no text to parse. Strengths: typed outputs, independent evaluation of questions, calibrated probabilities. Limits: built for well-scoped atomic questions; confidence only for some question types. No pricing on the page; undated, checked today.

Cited on:Jev (TypeSafe)

What it says: Founded January 2021 in San Francisco by former OpenAI employees including Dario and Daniela Amodei. Amazon invested $1.25 billion (September 2023), $2.75 billion (March 2024), and $4 billion (November 2024); Google invested $500 million (October 2023). Only the first 100,000 characters were read.

Cited on:Claude

What it says: Founded in 2022 by Polish entrepreneurs Piotr Dąbkowski and Mati Staniszewski; based in London with offices in New York City, Warsaw, and San Francisco. February 4, 2026 round of $500 million at an $11 billion valuation.

Cited on:Voice: ElevenLabs, Rime, and Kokoro

What it says: Founded September 4, 1998, in Menlo Park, California, by Larry Page and Sergey Brin; headquartered at the Googleplex, Mountain View; reorganized in August 2015 as a wholly owned subsidiary of Alphabet Inc. Only the first 100,000 characters were read.

Cited on:Google AI Ultra, Gemini, and Antigravity

What it says: Founded February 4, 2004, in Cambridge, Massachusetts, as TheFacebook, Inc., by Mark Zuckerberg, Eduardo Saverin, Dustin Moskovitz, Andrew McCollum, and Chris Hughes; headquartered in Menlo Park, California; renamed Meta Platforms, Inc. in 2021. Only the first 100,000 of 236,858 characters were read.

Cited on:Muse (Meta)

What it says: Founded December 2015 as the nonprofit OpenAI, Inc.; founders include Elon Musk, Sam Altman, Ilya Sutskever, Greg Brockman, and Trevor Blackwell; headquartered in San Francisco. Now OpenAI Group PBC, with the OpenAI Foundation appointing all board members. Microsoft stake given as 23.3% (infobox) and 27% (text). A 2026 round of $122 billion at a post-money $852 billion, dated both March and April. Only the first 100,000 characters were read.

Cited on:ChatGPT Pro and Codex

What it says: Founders Mikey Shulman, Georg Kucsko, Martin Camacho, and Keenan Freyberg, who all worked at Kensho; developer Suno, Inc. "of Cambridge, Massachusetts"; web app released December 20, 2023. Suno acquired Songkick from WMG as part of the November 2025 settlement. UMG and Sony remain plaintiffs; a second suit filed September 18, 2026 covers 60,202 recordings; Suno calls the claims "fundamentally flawed on both the facts and the law." Only the first 100,000 of 104,171 characters were read.

Cited on:Suno

What it says: Elon Musk founded xAI on March 9, 2023, with 11 researchers; all other co-founders had left by late March 2026. Headquarters: 1450 Page Mill Road, Palo Alto. Acquired X Corp. in March 2025; SpaceX acquired xAI in February 2026 (wholly owned subsidiary); renamed SpaceXAI in July 2026 and SpaceX Super Intelligence in October 2026. December 2024 round of $6 billion brought total funding to over $12 billion. Only the first 100,000 characters were read.

Cited on:Grok and Grok Build

What it says: Covers the Grok website and apps; names two subscriptions, SuperGrok and SuperGrok Heavy, without prices; describes a shared weekly usage system rolling out in June 2026. Undated page, checked today.

Cited on:Grok and Grok Build

What it says: Grok Build is "a powerful and extensible coding agent" usable through an interactive TUI, headlessly in scripts, or over the Agent Client Protocol; runs grok-4.7, which the page calls SpaceXSI's latest model. Does not say whether a subscription is required or give usage limits. Undated page, checked today.

Cited on:Grok and Grok Build

What it says: Grok Build is "a new coding agent that runs right from your terminal," an early beta open to "all SuperGrok and X Premium Plus subscribers"; supports plan-based review and headless mode. Lists no usage caps or pricing for Grok Build itself.

Cited on:Grok and Grok Build

What it says: Tiers: Free $0, Plus $8 (base), Premium $30 (4x Plus), Super $100 (15x Plus), Ultra $200 (50x Plus). Premium carries "the benefits of SuperGrok, Cursor Pro, and X Premium+, previously $90/month combined." Gives no SuperGrok price and does not mention SuperGrok Heavy.

Cited on:Grok and Grok Build, Home