Skip to Sidebar Skip to Content
Company tracker 2 of 10 · 10 companies, one lens

Anthropic

The maker of Claude, tracked through the same six lenses as every company on AI Adoption.
Its best model, Claude Opus 5.5, scores 168.8 on the Epoch Capabilities Index, the highest of the 10 tracked companies. Annualized revenue reached $65B by 31 Jul 2026, and its May 2026 round valued it at $965.0B. In March 2026, 3.0% of US adults said they had used Claude in the past week. It released 7 notable models in 2025, #4 of 10 by count.
Follow AnthropicCompare with the others
61 ledger entries22 charts name AnthropicLatest entry Sep 2026Checked 20 Oct 2026

Anthropic through six lenses

How the lenses work →
The same six lenses, in the same order, on every company page. Each shows one comparable number where a source measures several companies the same way, its rank, and when it last changed.
01Capability#1 of 10
How good are its best models?
168.8
Epoch Capabilities Index score of its best model, Claude Opus 5.5 (Sep 2026)
Anthropic168.8OpenAI168.8Google166.4Meta161.8
Epoch combines many benchmarks into one scale. 90% interval for this model: 165.2 to 173.9.
Text Arena, Mar 2026: 1,503, #1 of 8. Arena chart →
Independent measureChanged Sep 2026Epoch AI data →
02Output#4 of 10
How much does it ship?
7
Notable AI models released in 2025
OpenAI21Google15Alibaba11Anthropic7
13 notable models from 2014 to 2025 and 9 in 2026 so far. Model record →
Independent measureChanged 2025View chart →
03Money#1 of 9
How much does it earn?
$65B
Annualized revenue, 31 Jul 2026
Anthropic$65BOpenAI$50BZ.ai (Zhipu)$1.6BDeepSeek$1B
Annualized run rate; Epoch rates the report “Likely”. Dates differ by company; each is its latest report. Ranked among the 9 companies in Epoch AI's revenue data.
Valuation $965.0B after its May 2026 round; $139.4B raised in equity in all; about 3,000 staff (Aug 2026).
Reported figureChanged Jul 2026Epoch AI data →
04Compute#5 of 10
How much compute and energy goes into its AI?
3.4×10²⁵
FLOP: its largest training run with a published estimate, Claude 3.7 Sonnet (Feb 2025)
OpenAI9.4×10²⁶xAI5×10²⁶Meta5.2×10²⁵Google5×10²⁵Anthropic3.4×10²⁵
Largest estimated training cost: $26M (Claude 3.5 Sonnet), Epoch AI. Model record →
Ranked among the 10 tracked companies.
Reported compute spend
OpenAI spend, 2025$16.3BAnthropic spend, 2025$6.8B
R&D Inference. Epoch AI reports spend for these two labs only.
Had about 1.19M H100-equivalents available at the end of 2025, #3 of 5 labs. Epoch AI data →
Independent estimateChanged Feb 2025View chart →
05Use#6 of 7
Who uses it, and for what?
3.0%
of US adults said they used Claude in the past week, March 2026
OpenAI31.0%Google20.9%Microsoft10.5%Meta7.5%Anthropic3.0%
Epoch AI with Ipsos, a probability-based sample of 2,021 US adults; 49.2% used no AI service. Among employed adults in July 2026: 7.7%.
45% of sampled Claude.ai conversations were automation, Nov 2025. Chart →
From the company: how people use Claude
Jan 2025Nov 2025
Automation Augmentation
Independent measureChanged Mar 2026Epoch AI data →
06Safety & policy#2 of 9
How reliable is it, and how is it governed?
26%
hallucination rate of its most reliable model on AA-Omniscience, Claude 4.5 Haiku
xAI22%Anthropic26%Google50%DeepSeek82%
Lower is better. Ranked among the tracked companies; the test covers 26 models. One reliability test: safety frameworks, evaluations and regulatory actions are logged as entries.
Independent measureChanged 2026View chart →

The ledger

What gets an entry →
An entry goes in only if it changes a number or a position on one of the six lenses. If it moves nothing, it is not news here. Newest first.

  1. September 2026

    1. 22 Sep 2026CapabilityNewClaude Opus 5.5 scored 168.8 on the Epoch Capabilities Index, its best model#1 of the 10 tracked companies' best models; level with OpenAI at this precision. Epoch's 90% interval: 165.2 to 173.9.Independent measureEpoch AI, Benchmarking HubEpoch AI data →#1

  2. July 2026

    1. 31 Jul 2026MoneyNewAnnualized revenue reached $65B, from $47B on 15 May 2026OpenAI's latest estimate at that date was $30B (15 Jul 2026). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$18B
    2. Jul 2026UseNew7.7% of employed US adults said they had used Claude in the past weekProbability-based sample of 1,103 employed US adults, asked about work and personal use.Independent measureEpoch AI and IpsosEpoch AI data →

  3. May 2026

    1. 28 May 2026MoneyNewClosed a funding round at a $965.0B post-money valuation, raising $65.0B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$965.0B
    2. 15 May 2026MoneyNewAnnualized revenue reached $47B, from $30B on 6 Apr 2026OpenAI's latest estimate at that date was $25B (28 Feb 2026). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$17B

  4. April 2026

    1. 24 Apr 2026MoneyNewClosed a funding round at a $380.0B post-money valuation, raising $10.0B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$380.0B
    2. 20 Apr 2026MoneyNewClosed a funding round at a $380.0B post-money valuation, raising $5.0B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$380.0B
    3. 6 Apr 2026MoneyNewAnnualized revenue reached $30B, from $19B on 3 Mar 2026OpenAI's latest estimate at that date was $25B (28 Feb 2026). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$11B

  5. March 2026

    1. Mar 2026UseNew3.0% of US adults said they had used Claude in the past week#6 of the 7 services asked about that belong to tracked companies. 49.2% had used no AI service. Probability-based sample of 2,021 US adults.Independent measureEpoch AI and IpsosEpoch AI data →#6

  6. 2026 tests

    1. 2026 testSafety & policyNewIts most reliable model on AA-Omniscience, Claude 4.5 Haiku, had a 26% hallucination rate#2 of 9 tracked companies; xAI's best was 22%. Rates across the 26 models tested ran from 22% to 94%.Independent measureArtificial AnalysisView chart →#2 of 9

  7. March 2026

    1. 3 Mar 2026MoneyNewAnnualized revenue reached $19B, from $14B on 12 Feb 2026OpenAI's latest estimate at that date was $25B (28 Feb 2026). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$5B
    2. Mar 2026CapabilityNewStill first with 1,503; the top four labs sat within 22 Arena pointsxAI 1,495, Google 1,494, OpenAI 1,481. In May 2023 the leader was 205 points clear.Independent measureArenaView chart →#1

  8. February 2026

    1. Feb 2026CapabilityNewClaude 4.5 Opus led SWE-bench Verified at 76.8%The top 10 models sat within 6 points; Kimi K2.5 was tenth at 70.8%.Independent measureSWE-bench LeaderboardView chart →#1
    2. 12 Feb 2026MoneyNewAnnualized revenue reached $14B, from $9B on 31 Dec 2025OpenAI's latest estimate at that date was $21.4B (31 Dec 2025). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$5B
    3. 12 Feb 2026MoneyNewClosed a funding round at a $380.0B post-money valuation, raising $30.0B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$380.0B
    4. Feb 2026CapabilityNewTook first place on the Text Arena with 1,504 pointsGoogle was second at 1,500. First to fourth spanned 23 points.Independent measureArenaView chart →#3 → #1

  9. Early 2026 leaderboards

    1. Early 2026CapabilityNewClaude Opus 4.6 solved 93% of Cybench security challenges unguidedThe best result in 2024 was 17.5%, from Claude 3.5 Sonnet.Independent measureCybench LeaderboardView chart →#1
    2. Early 2026CapabilityNewClaude Opus 4.6 led Vibe Code Bench at 57.6%Building complete web apps from a prompt. The lowest of eight models scored 31.5%.Independent measureVals.aiView chart →#1
    3. Early 2026CapabilityNewClaude Opus 4.5 led τ-bench tool use at 70.2%, 0.3 points ahead of GPT-5.2No model shown passed 71%.Independent measureτ-bench LeaderboardView chart →#1
    4. Early 2026CapabilityNewClaude-Opus-4.5 led the Berkeley Function Calling Leaderboard at 77.5%4.2 points ahead of the next entry, Claude-Sonnet-4.5.Independent measureBerkeley Function Calling LeaderboardView chart →#1
    5. Early 2026CapabilityNewClaude Sonnet 4.6 led Finance Agent at 63.3%537 entry-level analyst tasks. No model reached two-thirds.Independent measureVals.aiView chart →#1
    6. Early 2026CapabilityNewClaude Sonnet 4.6 led TaxEval v2 at 77.1%The top 15 models were within 2.9 points of each other.Independent measureVals.aiView chart →#1

  10. December 2025

    1. 31 Dec 2025MoneyNewAnnualized revenue reached $9B, from $7B on 21 Oct 2025OpenAI's latest estimate at that date was $21.4B (31 Dec 2025). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$2B

  11. End of 2025

    1. End of 2025ComputeNewHad compute equal to about 1.19 million H100s available at the end of 2025Owned or rented. Epoch's 90% range: 0.84M to 1.72M. #3 of 5 labs.Independent estimateEpoch AI, AI Chip UsersEpoch AI data →1.19M

  12. 2025, full year

    1. 2025ComputeNewReported compute spend was $6.8B in 2025, up from $2.5B$4.1B on R&D and $2.7B on inference. OpenAI's was $16.3B.Independent estimateEpoch AIView chart →2.7×
    2. 2025OutputNewReleased 7 notable AI models in 2025, #4 by countOpenAI 21, Google 15, Alibaba 11. Epoch AI counts a model as notable if it was state of the art, highly cited or historically significant.Independent measureEpoch AIView chart →#4
    3. 2025CapabilityNewClaude Opus 4.5 set the best OSWorld score of 2025, 66.3%Computer science students complete 72.4% of the tasks.Independent measureEpoch AIView chart →#1

  13. November 2025

    1. Nov 2025CapabilityNewFell to #3 of 8 on the Text Arena, 26 points behind GoogleGoogle led at 1,492. First to fourth spanned 31 points.Independent measureArenaView chart →#2 → #3
    2. Nov 2025UseNewAugmentation moved back ahead, 52% to 45%From 49% and 47% in Aug 2025.Company dataAnthropic Economic IndexView chart →-4 pts
    3. Nov 2025UseNewEducation tasks reached 15% of sampled Claude conversationsFrom 9.6% in Jan 2025. Computer and maths tasks stayed the largest group, at 34%.Company dataAnthropic Economic IndexView chart →+5.4 pts

  14. October 2025

    1. 21 Oct 2025MoneyNewAnnualized revenue reached $7B, from $5B on 29 Jul 2025OpenAI's latest estimate at that date was $13B (1 Aug 2025). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$2B

  15. September 2025

    1. 2 Sep 2025MoneyNewClosed a funding round at a $183.0B post-money valuation, raising $13.0B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$183.0B

  16. August 2025

    1. Aug 2025CapabilityNewRose to #2 of 8 on the Text Arena, 3 points behind GoogleGoogle led at 1,461. First to fourth spanned 30 points.Independent measureArenaView chart →#5 → #2
    2. Aug 2025UseNewAutomation overtook augmentation, 49% to 47%From 42% and 55% in Mar 2025.Company dataAnthropic Economic IndexView chart →+7 pts

  17. July 2025

    1. 29 Jul 2025MoneyNewAnnualized revenue reached $5B, from $4B on 1 Jul 2025OpenAI's latest estimate at that date was $10B (9 Jun 2025). Annual recurring revenue (ARR); Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$1B
    2. Jul 2025CapabilityNewFell to #5 of 8 on the Text Arena, 59 points behind OpenAIOpenAI led at 1,482. First to fourth spanned 50 points.Independent measureArenaView chart →#4 → #5
    3. 1 Jul 2025MoneyNewAnnualized revenue reached $4B, from $3B on 30 May 2025OpenAI's latest estimate at that date was $10B (9 Jun 2025). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$1B

  18. June 2025

    1. Jun 2025CapabilityNewFell to #4 of 8 on the Text Arena, 52 points behind GoogleGoogle led at 1,472. First to fourth spanned 52 points.Independent measureArenaView chart →#3 → #4

  19. May 2025

    1. 30 May 2025MoneyNewAnnualized revenue reached $3B, from $2B on 31 Mar 2025OpenAI's latest estimate at that date was $5.5B (31 Dec 2024). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$1B
    2. May 2025CapabilityNewRose to #3 of 8 on the Text Arena, 69 points behind GoogleGoogle led at 1,492. First to fourth spanned 80 points.Independent measureArenaView chart →#4 → #3

  20. April 2025

    1. Apr 2025CapabilityNewFell to #4 of 8 on the Text Arena, 70 points behind OpenAIOpenAI led at 1,461. First to fourth spanned 70 points.Independent measureArenaView chart →#3 → #4

  21. March 2025

    1. 31 Mar 2025MoneyNewAnnualized revenue reached $2B, from $1.4B on 1 Mar 2025OpenAI's latest estimate at that date was $5.5B (31 Dec 2024). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$600M
    2. 3 Mar 2025MoneyNewClosed a funding round at a $61.5B post-money valuation, raising $3.5B in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$61.5B
    3. 1 Mar 2025MoneyNewAnnualized revenue reached $1.4B, from $1B on 31 Dec 2024OpenAI's latest estimate at that date was $5.5B (31 Dec 2024). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$400M
    4. Mar 2025UseNewAutomation was 42% of sampled conversations, augmentation 55%From 41% and 55% in Jan 2025.Company dataAnthropic Economic IndexView chart →+1 pts

  22. February 2025

    1. 24 Feb 2025OutputNewClaude 3.7 Sonnet released, trained with an estimated 3.4×10²⁵ FLOPIndependent estimateEpoch AIView chart →3.4×10²⁵

  23. January 2025

    1. Jan 2025CapabilityNewFell to #3 of 8 on the Text Arena, 40 points behind OpenAIOpenAI led at 1,405. First to fourth spanned 56 points.Independent measureArenaView chart →#2 → #3
    2. Jan 2025UseNewFirst reading: automation was 41% of sampled Claude.ai conversations, augmentation 55%Augmentation means the person works with Claude; automation means Claude does the task.Company dataAnthropic Economic IndexView chart →

  24. December 2024

    1. 31 Dec 2024MoneyNewAnnualized revenue reached $1B, from $87M on 1 Jan 2024OpenAI's latest estimate at that date was $5.5B (31 Dec 2024). Annualized run rate; Epoch rates the report “Likely”.Reported figureEpoch AI, AI CompaniesEpoch AI data →+$913M

  25. 2024, full year

    1. 2024ComputeNewReported compute spend was $2.5B in 2024, up from $285M$1.5B on R&D. OpenAI's was $5.8B.Independent estimateEpoch AIView chart →8.8×

  26. October 2024

    1. Oct 2024CapabilityNewRose to #2 of 8 on the Text Arena, 20 points behind OpenAIOpenAI led at 1,384. First to fourth spanned 50 points.Independent measureArenaView chart →#3 → #2

  27. September 2024

    1. Sep 2024CapabilityNewFell to #3 of 8 on the Text Arena, 50 points behind OpenAIOpenAI led at 1,387. First to fourth spanned 53 points.Independent measureArenaView chart →#2 → #3

  28. June 2024

    1. 20 Jun 2024OutputNewClaude 3.5 Sonnet released, trained with an estimated 2.7×10²⁵ FLOPAbout 7× its previous model in this data, Claude 2 (Jul 2023).Independent estimateEpoch AIView chart →2.7×10²⁵

  29. April 2024

    1. Apr 2024CapabilityNewFell to #2 of 7 on the Text Arena, 15 points behind OpenAIOpenAI led at 1,338. First to fourth spanned 91 points.Independent measureArenaView chart →#1 → #2

  30. March 2024

    1. 4 Mar 2024OutputNewClaude 3 Opus released, marked a frontier model in Epoch AI's databaseIndependent estimateEpoch AIModel record →
    2. Mar 2024CapabilityNewTook first place on the Text Arena with 1,333 points, its first month in this seriesOpenAI was second at 1,325. First to fourth spanned 95 points.Independent measureArenaView chart →#1

  31. January 2024

    1. 11 Jan 2024MoneyNewClosed a funding round at a $18.4B post-money valuation, raising $750M in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$18.4B
    2. 1 Jan 2024MoneyNewAnnualized revenue was estimated at $87M, against $100M on 1 Oct 2023OpenAI's latest estimate at that date was $2B (31 Dec 2023). Annualized run rate; Epoch rates the report “Confident”.Reported figureEpoch AI, AI CompaniesEpoch AI data →−$13M

  32. 2023, full year

    1. 2023ComputeNewReported compute spend was $285M in 2023Independent estimateEpoch AIView chart →

  33. July 2023

    1. 11 Jul 2023OutputNewClaude 2 released, trained with an estimated 3.9×10²⁴ FLOPIndependent estimateEpoch AIView chart →3.9×10²⁴

  34. May 2023

    1. 23 May 2023MoneyNewClosed a funding round at a $4.5B post-money valuation, raising $450M in equityReported figureEpoch AI, AI CompaniesEpoch AI data →$4.5B

Anthropic's model record

Download the CSV →
Every notable model Epoch AI has logged for Anthropic: 22 in all, 13 from 2014 to 2025 and 9 in 2026 so far. Training compute, cost and hardware are Epoch's estimates, shown only where it has one.
Notable models, 2014–202513#7 of 10 companies · Epoch AI's count, as in chart 1.1.7
Frontier models3#5 of 10 companies · marked frontier in Epoch AI's database
Open weights0%0 of 22 notable models since 2023
Largest estimated training cost$26M#5 of 10 companies · Claude 3.5 Sonnet (Jun 2024), 2023 dollars
Largest training cluster—No estimate in Epoch AI's database
Notable models per year* 2026 to 19 Oct 2026
2014: 020142015: 02016: 020162017: 02018: 020182019: 02020: 020202021: 02022: 020222023: 332024: 3320242025: 772026: 992026*
Notable models since 2023, newest first22 models
ReleasedModelAccessTraining computeEst. costTraining hardware
28 Sep 2026Claude Sonnet 5.5API or hosted access———
22 Sep 2026Claude Opus 5.5API or hosted access———
1 Sep 2026Claude Fable 5.1API or hosted access———
24 Jul 2026Claude Opus 5API or hosted access———
9 Jun 2026Claude Mythos 5Unreleased———
9 Jun 2026Claude Fable 5API or hosted access———
28 May 2026Claude Opus 4.8API or hosted access———
16 Apr 2026Claude Opus 4.7API or hosted access———
7 Apr 2026Claude Mythos PreviewAPI or hosted access———
24 Nov 2025Claude Opus 4.5API or hosted access———
Show all 22 models since 2023
15 Oct 2025Claude Haiku 4.5API or hosted access———
29 Sep 2025Claude Sonnet 4.5API or hosted access———
5 Aug 2025Claude Opus 4.1API or hosted access———
22 May 2025Claude Sonnet 4API or hosted access———
22 May 2025Claude Opus 4API or hosted access———
24 Feb 2025Claude 3.7 SonnetAPI or hosted access3.4×10²⁵——
20 Jun 2024Claude 3.5 Sonnet frontierAPI or hosted access2.7×10²⁵$26M—
4 Mar 2024Claude 3 SonnetAPI or hosted access———
4 Mar 2024Claude 3 Opus frontierAPI or hosted access———
21 Nov 2023Claude 2.1API or hosted access———
11 Jul 2023Claude 2 frontierAPI or hosted access3.9×10²⁴$5M—
14 Mar 2023ClaudeAPI or hosted access———
Data: Epoch AI, Data on AI Models, snapshot 19 Oct 2026, CC BY 4.0. Organisations are grouped into companies as in chart 1.1.7 (for example, Google DeepMind and Google Brain count as Google). Anthropic's models as CSV.

Anthropic beside the others

All 10 companies →
The same rows for every company. A dash means no source measures it for that company yet: gaps are shown, never filled with a guess. The small number is the rank among the companies in the full table that the source covers.
LensAnthropicOpenAIGooglexAIMetaMicrosoftDeepSeekAlibabaMistral AINvidia
CapabilityEpoch Capabilities Index, its best model · data168.8#1168.8#1166.4#3156.6#6161.8#4130.5#10155.9#7157.4#5150.0#8146.0#9
OutputNotable AI models released in 2025 · chart7#421#115#25#54#60#104#611#31#93#8
MoneyAnnualized revenue, latest report · data$65B#1Jul 2026$50B#2Sep 2026—$428M#5Sep 2025——$1B#3Sep 2026—$1B#3Sep 2026—
ComputeLargest training run with a published estimate, FLOP · chart3.4×10²⁵#5Claude 3.7 Sonnet, Feb 20259.4×10²⁶#1GPT-6 Astra, Sep 20265×10²⁵#4Gemini 1.0 Ultra, Dec 20235×10²⁶#2Grok 4, Jul 20255.2×10²⁵#3Llama 4 Behemoth (preview), Apr 20258.6×10²³#9Megatron-Turing NLG 530B, Oct 2021 (with Nvidia)9.7×10²⁴#8DeepSeek-V4-Pro, Apr 20261.5×10²⁵#7Qwen3-Max, Sep 20257.7×10²³#10Mixtral 8x7B, Dec 20231.8×10²⁵#6Nemotron-4 340B, Jun 2024
UseUS adults who used its assistant in the past week, Mar 2026 · data3.0%#631.0%#120.9%#24.5%#57.5%#410.5%#30.6%#7———
Safety & policyLowest hallucination rate of its models (lower is better) · chart26%#289%#850%#322%#187%#7—82%#489%#884%#683%#5
More measures · Epoch AI data
Text Arenabest score, Mar 2026 · data1,503#11,481#41,494#31,495#21,335#8—1,424#61,449#51,416#7—
Valuationlatest closed round · data$965.0B#1$852.0B#2—$230.0B#3——$50.0B#4—$24.0B#5—
Equity raisedin all · data$139.4B#2$182.8B#1—$37.0B#3——$7.5B#4—$6.5B#5—
Stafflatest report · data3,000#44,500#26,000#1300#63,400#3—300#6—350#5—
AI chips ownedH100-equivalents · data——5.47M#10.55M#41.95M#33.32M#2————
Compute availableH100-equivalents, end of 2025 · data1.19M#31.74M#11.58M#20.63M#51.00M#4—————
Weekly use at workemployed US adults, Jul 2026 · data7.7%#533.7%#124.2%#23.3%#69.8%#414.8%#3————
Model record · Epoch AI estimates
Notable models, 2014–2025Epoch AI's count, as in chart 1.1.713#761#3194#17#887#242#47#827#66#1030#5
Frontier modelsmarked frontier in Epoch AI's database3#515#233#13#55#33#50#80#80#84#4
Open weightsshare of its notable models since 20230%0 of 224%2 of 5714%8 of 5711%1 of 973%19 of 2633%2 of 6100%10 of 1075%24 of 3267%4 of 683%15 of 18
Largest estimated training cost2023 US dollars, among models Epoch has costed$26M#5Claude 3.5 Sonnet$366M#2GPT-4.5$31M#4Gemini 1.0 Ultra$388M#1Grok 4$53M#3Llama 3.1-405B$4M#9Megatron-Turing NLG 530B$7M#8DeepSeek-R1 (May 2025)$109K#10PLUG$14M#7Mistral Large$21M#6Nemotron-4 340B
Largest training clusterchips for one model, where Epoch records it—100,000#2GPT-6 Astra57,000#3Gemini 1.0 Ultra200,000#1Grok 432,000#4Llama 4 Behemoth (preview)4,480#6Megatron-Turing NLG 530B2,048#7DeepSeek-V3 (Mar 2025)512#8CogView—6,144#5Nemotron-4 340B
Charts on AI Adoptionthat name the company2233238404714

Every chart that names Anthropic

All charts →
22 charts across 6 hubs name Anthropic or one of its models. Each has its source and a CSV of its data.
AI Benchmarks13 charts
  1. 1,503Claude Opus 4.6 tops the Text Arena at 1,503, and the top 15 models span 46 points
  2. 31.3%The best model solved 31.3% of FrontierMath's hardest problems in 2025, up from 0% in 2024
  3. 76.8%Claude 4.5 Opus led SWE-bench Verified at 76.8% in February 2026; the top 10 sat within 6 points
  4. 57.6%Claude Opus 4.6 led Vibe Code Bench at 57.6% in early 2026, 26 points above the lowest of eight models
  5. 77.5%Claude-Opus-4.5 leads the Berkeley Function Calling Leaderboard at 77.5%, 4.2 points clear of the next
  6. 74.5%The best model scored 74.5% on GAIA in September 2025, 17.5 points short of the 92% human baseline
  7. 66.3%The best AI agent completed 66.3% of OSWorld computer tasks in 2025, against 72.4% for people
  8. 93%AI agents solved 93% of Cybench security challenges unguided in early 2026, up from 17.5% in 2024
  9. 70.2%Claude Opus 4.5 led τ-bench's tool-use tasks at 70.2% in early 2026; the top seven span 7.3 points
  10. 80%GPT-o1 answered 80% of CalendarQA date questions correctly; no other model of seven passed 47%
  11. 66%No model reached the 74.4% human baseline on Video-MMMU; the best, Keye-VL-1.5-8B, scored 66%
  12. 2.9 ptsThe top 15 models on TaxEval v2 scored within 2.9 percentage points of each other in early 2026
  13. 63.3%Claude Sonnet 4.6 led Finance Agent at 63.3% in early 2026; no model reached two-thirds
AI Models3 charts
  1. 16K2 Think and Olmo 3 32B Think top the Openness Index at 16 of 100; no other model scores on pre-training data
  2. 21OpenAI released the most notable AI models in 2025: 21, ahead of Google's 15
  3. 1,503Anthropic, xAI, Google and OpenAI's best models were within 22 Arena points in March 2026
AI Investment2 charts
  1. $65BAnthropic's annualized revenue reached $65B by July 2026; OpenAI's reached $50B by September
  2. $16.3BOpenAI's reported compute spend nearly tripled to $16.3B in 2025; Anthropic's reached $6.8B
AI Adoption2 charts
  1. 15%Education tasks rose to 15% of sampled Claude conversations by November 2025, from 9.6% in January
  2. 45%Automation rose to 45% of Claude.ai conversations by November 2025, after briefly overtaking augmentation
AI Research1 chart
  1. 0.53Ai2's Asta v0 agent scored 0.53 on AstaBench at $3.40 a problem; no other agent tested passed 0.44
AI Ethics & Safety1 chart
  1. 22%Grok 4.20 Beta had the lowest hallucination rate of 26 models on AA-Omniscience, 22%; gpt-oss-20B had 94%

The other companies, same lens

AI Companies →
OpenAI168.8Epoch Capabilities Index65 ledger entriesOpen tracker →Google166.4Epoch Capabilities Index48 ledger entriesOpen tracker →xAI156.6Epoch Capabilities Index37 ledger entriesOpen tracker →Meta161.8Epoch Capabilities Index33 ledger entriesOpen tracker →Microsoft130.5Epoch Capabilities Index4 ledger entriesOpen tracker →DeepSeek155.9Epoch Capabilities Index35 ledger entriesOpen tracker →Alibaba157.4Epoch Capabilities Index33 ledger entriesOpen tracker →Mistral AI150.0Epoch Capabilities Index26 ledger entriesOpen tracker →Nvidia146.0Epoch Capabilities Index7 ledger entriesOpen tracker →

How the ledger works

  1. One lens for everyone. Six lenses, same order, same definitions, for every company. No company gets an extra lens for a product it wants to talk about.
  2. Changes, not news. An entry must move a number or a rank on a lens. Launch events, interviews and hires do not count unless they do.
  3. Evidence tier on every line. Independent measure, independent estimate or company data. A company's own claims are replaced when an independent score arrives.
  4. Gaps stay visible. If no source measures a lens for a company, the page says so.
Disclosure: AI Adoption uses AI tools in its research and editorial work, including products from some of the companies tracked here. No company has any say in its page or its ledger, and the same rules apply to every company. Editorial standards · Data sources
Copied with its source. Want the next number like this in your inbox?Get Friday's numbers