Skip to Sidebar Skip to Content
Company tracker 8 of 10 · 10 companies, one lens

Alibaba

Alibaba, maker of the Qwen models, tracked through the same six lenses as every company on AI Adoption.
Its best model, Qwen 3.8 Max, scores 157.4 on the Epoch Capabilities Index, #5 of 10, 11.4 behind Anthropic. It released 11 notable models in 2025, #3 of 10 by count. Its largest training run with a published estimate, Qwen3-Max, used 1.5×10²⁵ FLOP. On one reliability test, AA-Omniscience, its best model's hallucination rate was 89%, #8 of 9.
Follow AlibabaCompare with the others
33 ledger entries7 charts name AlibabaLatest entry Aug 2026Checked 20 Oct 2026

Alibaba 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#5 of 10
How good are its best models?
157.4
Epoch Capabilities Index score of its best model, Qwen 3.8 Max (Aug 2026)
Anthropic168.8OpenAI168.8Google166.4Meta161.8Alibaba157.4
Epoch combines many benchmarks into one scale. 90% interval for this model: 155.3 to 160.3.
Text Arena, Mar 2026: 1,449, #5 of 8. Arena chart →
Independent measureChanged Aug 2026Epoch AI data →
02Output#3 of 10
How much does it ship?
11
Notable AI models released in 2025
OpenAI21Google15Alibaba11Anthropic7
27 notable models from 2014 to 2025 and 10 in 2026 so far. Model record →
Independent measureChanged 2025View chart →
03Money
How much does it earn?
Not measured yet
No AI revenue figure for Alibaba in Epoch AI's company data
Epoch AI tracks revenue for AI-first companies such as OpenAI, Anthropic, Mistral AI, DeepSeek.
How a lens gets a number →
04Compute#7 of 10
How much compute and energy goes into its AI?
1.5×10²⁵
FLOP: its largest training run with a published estimate, Qwen3-Max (Sep 2025)
OpenAI9.4×10²⁶xAI5×10²⁶Meta5.2×10²⁵Google5×10²⁵Alibaba1.5×10²⁵
Largest estimated training cost: $109K (PLUG), Epoch AI. Model record →
Ranked among the 10 tracked companies. Only runs with a published estimate count, so a company that discloses less can rank lower.
Independent estimateChanged Sep 2025View chart →
05Use
Who uses it, and for what?
Not measured yet
No Alibaba service in the Epoch AI / Ipsos poll
The poll asks about ChatGPT, Claude, Gemini, Copilot, Grok, Meta AI and DeepSeek.
How a lens gets a number →
06Safety & policy#8 of 9
How reliable is it, and how is it governed?
89%
hallucination rate of its most reliable model on AA-Omniscience, Qwen3.5 397B A17B
xAI22%Anthropic26%Google50%DeepSeek82%Alibaba89%
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. August 2026

    1. 2 Aug 2026CapabilityNewQwen 3.8 Max scored 157.4 on the Epoch Capabilities Index, its best model#5 of the 10 tracked companies' best models; Anthropic's best scored 168.8. Epoch's 90% interval: 155.3 to 160.3.Independent measureEpoch AI, Benchmarking HubEpoch AI data →#5

  2. March 2026

    1. Mar 2026CapabilityNewQwen 3-VL 235B Instruct read 39.4% of ClockBench clocks correctly, second of 15 modelsGPT-5.4 High led at 50.6%; humans manage about 90%.Independent measureClockBench LeaderboardView chart →#2

  3. 2026 tests

    1. 2026 testSafety & policyNewIts most reliable model on AA-Omniscience, Qwen3.5 397B A17B, had a 89% hallucination rate#8 of 9 tracked companies; xAI's best was 22%. Rates across the 26 models tested ran from 22% to 94%.Independent measureArtificial AnalysisView chart →#8 of 9

  4. March 2026

    1. Mar 2026CapabilityNewStill #5 of 8 with 1,449; the top four labs sat within 22 Arena pointsAnthropic 1,503, xAI 1,495, Google 1,494, OpenAI 1,481. In May 2023 the leader was 205 points clear.Independent measureArenaView chart →#5

  5. 2025, full year

    1. 2025OutputNewReleased 11 notable AI models in 2025, #3 by countOpenAI 21, Google 15. Epoch AI counts a model as notable if it was state of the art, highly cited or historically significant.Independent measureEpoch AIView chart →#3

  6. November 2025

    1. Nov 2025CapabilityNewFell to #5 of 8 on the Text Arena, 59 points behind GoogleGoogle led at 1,492. First to fourth spanned 31 points.Independent measureArenaView chart →#4 → #5

  7. September 2025

    1. 5 Sep 2025OutputNewQwen3-Max released, trained with an estimated 1.5×10²⁵ FLOPAbout 3× its previous model in this data, Qwen3-235B-A22B (Apr 2025).Independent estimateEpoch AIView chart →1.5×10²⁵
    2. Sep 2025CapabilityNewRose to #4 of 8 on the Text Arena, 17 points behind GoogleGoogle led at 1,453. First to fourth spanned 17 points.Independent measureArenaView chart →#5 → #4

  8. August 2025

    1. Aug 2025CapabilityNewFell to #5 of 8 on the Text Arena, 31 points behind GoogleGoogle led at 1,461. First to fourth spanned 30 points.Independent measureArenaView chart →#3 → #5

  9. July 2025

    1. Jul 2025CapabilityNewRose to #3 of 8 on the Text Arena, 44 points behind OpenAIOpenAI led at 1,482. First to fourth spanned 50 points.Independent measureArenaView chart →#6 → #3

  10. 2025 studies

    1. 2025 studySafety & policyNewQwen-72B was the best open-source model on ResponsibleRobotBench, at 0.35 safe successGPT-4o led the five models tested at 0.64.Independent measureZhang et al.View chart →
    2. 2025 studyCapabilityNewQwen2-VL-7B read none of 62 analogue clocks exactly right on ClockQAThe best of seven models, Gemini-2.0, managed 22.6%.Independent measureSaxena et al.View chart →

  11. June 2025

    1. Jun 2025CapabilityNewRose to #6 of 8 on the Text Arena, 73 points behind GoogleGoogle led at 1,472. First to fourth spanned 52 points.Independent measureArenaView chart →#7 → #6

  12. May 2025

    1. May 2025CapabilityNewFell to #7 of 8 on the Text Arena, 120 points behind GoogleGoogle led at 1,492. First to fourth spanned 80 points.Independent measureArenaView chart →#5 → #7

  13. April 2025

    1. 29 Apr 2025OutputNewQwen3-235B-A22B released, trained with an estimated 4.8×10²⁴ FLOPLess training compute than Qwen2.5-72B (7.8×10²⁴).Independent estimateEpoch AIView chart →4.8×10²⁴
    2. Apr 2025CapabilityNewRose to #5 of 8 on the Text Arena, 91 points behind OpenAIOpenAI led at 1,461. First to fourth spanned 70 points.Independent measureArenaView chart →#6 → #5

  14. March 2025

    1. Mar 2025CapabilityNewFell to #6 of 8 on the Text Arena, 72 points behind OpenAIOpenAI led at 1,441. First to fourth spanned 46 points.Independent measureArenaView chart →#5 → #6

  15. February 2025

    1. Feb 2025CapabilityNewRose to #5 of 8 on the Text Arena, 41 points behind OpenAIOpenAI led at 1,405. First to fourth spanned 33 points.Independent measureArenaView chart →#7 → #5

  16. December 2024

    1. Dec 2024CapabilityNewFell to #7 of 8 on the Text Arena, 93 points behind OpenAIOpenAI led at 1,407. First to fourth spanned 56 points.Independent measureArenaView chart →#6 → #7

  17. October 2024

    1. Oct 2024CapabilityNewRose to #6 of 8 on the Text Arena, 74 points behind OpenAIOpenAI led at 1,384. First to fourth spanned 50 points.Independent measureArenaView chart →#7 → #6

  18. September 2024

    1. 19 Sep 2024OutputNewQwen2.5-72B released, trained with an estimated 7.8×10²⁴ FLOPAbout 3× its previous model in this data, Qwen2-72B (Jun 2024).Independent estimateEpoch AIView chart →7.8×10²⁴

  19. August 2024

    1. Aug 2024CapabilityNewFell to #7 of 8 on the Text Arena, 80 points behind OpenAIOpenAI led at 1,341. First to fourth spanned 10 points.Independent measureArenaView chart →#6 → #7

  20. July 2024

    1. Jul 2024CapabilityNewFell to #6 of 7 on the Text Arena, 79 points behind OpenAIOpenAI led at 1,341. First to fourth spanned 14 points.Independent measureArenaView chart →#5 → #6

  21. June 2024

    1. 7 Jun 2024OutputNewQwen2-72B released, trained with an estimated 3×10²⁴ FLOPAbout 2× its previous model in this data, Qwen-72B (Nov 2023).Independent estimateEpoch AIView chart →3×10²⁴
    2. Jun 2024CapabilityNewRose to #5 of 7 on the Text Arena, 81 points behind OpenAIOpenAI led at 1,341. First to fourth spanned 71 points.Independent measureArenaView chart →#6 → #5

  22. May 2024

    1. May 2024CapabilityNewFell to #6 of 7 on the Text Arena, 103 points behind OpenAIOpenAI led at 1,340. First to fourth spanned 70 points.Independent measureArenaView chart →#5 → #6

  23. April 2024

    1. Apr 2024CapabilityNewFell to #5 of 7 on the Text Arena, 103 points behind OpenAIOpenAI led at 1,338. First to fourth spanned 91 points.Independent measureArenaView chart →#4 → #5

  24. March 2024

    1. Mar 2024CapabilityNewFell to #4 of 7 on the Text Arena, 95 points behind AnthropicAnthropic led at 1,333. First to fourth spanned 95 points.Independent measureArenaView chart →#3 → #4

  25. February 2024

    1. Feb 2024CapabilityNewRose to #3 of 6 on the Text Arena, 95 points behind OpenAIOpenAI led at 1,329. First to fourth spanned 113 points.Independent measureArenaView chart →#6 → #3

  26. January 2024

    1. Jan 2024CapabilityNewFell to #6 of 6 on the Text Arena, 193 points behind OpenAIOpenAI led at 1,334. First to fourth spanned 145 points.Independent measureArenaView chart →#4 → #6

  27. December 2023

    1. Dec 2023CapabilityNewFell to #4 of 5 on the Text Arena, 163 points behind OpenAIOpenAI led at 1,312. First to fourth spanned 163 points.Independent measureArenaView chart →#2 → #4

  28. November 2023

    1. 30 Nov 2023OutputNewQwen-72B released, trained with an estimated 1.3×10²⁴ FLOPIndependent estimateEpoch AIView chart →1.3×10²⁴

  29. October 2023

    1. Oct 2023CapabilityNewPlaced #2 of 5 on the Text Arena with 1,154 points, its first month in this seriesOpenAI led at 1,314. First to fourth spanned 184 points.Independent measureArenaView chart →#2

Alibaba's model record

Download the CSV →
Every notable model Epoch AI has logged for Alibaba: 37 in all, 27 from 2014 to 2025 and 10 in 2026 so far. Training compute, cost and hardware are Epoch's estimates, shown only where it has one.
Notable models, 2014–202527#6 of 10 companies · Epoch AI's count, as in chart 1.1.7
Frontier models0#8 of 10 companies · marked frontier in Epoch AI's database
Open weights75%24 of 32 notable models since 2023
Largest estimated training cost$109K#10 of 10 companies · PLUG (Apr 2021), 2023 dollars
Largest training cluster512#8 of 8 companies · NVIDIA Tesla V100 DGXS 16 GB · CogView
Notable models per year* 2026 to 19 Oct 2026
2014: 020142015: 02016: 020162017: 02018: 020182019: 02020: 020202021: 442022: 1120222023: 552024: 6620242025: 11112026: 10102026*
Notable models since 2023, newest first32 models
ReleasedModelAccessTraining computeEst. costTraining hardware
8 Aug 2026Qwen3.8-2.4T-A95BOpen weights———
19 Jul 2026Qwen 3.8 MaxAPI or hosted access———
3 Jun 2026Qwen3.7-PlusAPI or hosted access———
1 Apr 2026Qwen 3.6 PlusAPI or hosted access———
29 Mar 2026Qwen3.5-Omni-PlusAPI or hosted access———
29 Mar 2026Qwen3.5-Omni-FlashAPI or hosted access———
24 Feb 2026Qwen3.5-122B-A10BOpen weights———
15 Feb 2026Qwen3.5 397B-A17BOpen weights———
2 Feb 2026Qwen3-Coder-NextOpen weights———
25 Jan 2026Qwen3-Max-ThinkingAPI or hosted access———
Show all 32 models since 2023
28 Oct 2025Tongyi DeepResearchOpen weights———
22 Sep 2025Qwen3-Omni-30B-A3BOpen weights3.6×10²²——
16 Sep 2025AgentFounder-30BOpen weights6.5×10²³——
5 Sep 2025Qwen3-MaxAPI or hosted access1.5×10²⁵——
4 Aug 2025Qwen ImageOpen weights———
25 Jul 2025Qwen3-235B-A22B-Thinking (Jul 2025)Open weights4.8×10²⁴——
25 Jul 2025Qwen3-235B-A22B (Jul 2025)Open weights4.8×10²⁴——
22 Jul 2025Qwen3-Coder-480B-A35BOpen weights1.6×10²⁴——
5 Jun 2025Qwen3 EmbeddingOpen weights———
29 Apr 2025Qwen3-235B-A22BOpen weights4.8×10²⁴——
6 Mar 2025QwQ-32BOpen weights3.5×10²⁴——
19 Sep 2024Qwen2.5-72BOpen weights7.8×10²⁴——
19 Sep 2024Qwen2.5 Instruct (72B)Open weights7.9×10²⁴——
17 Sep 2024Qwen2.5-32BOpen weights3.5×10²⁴——
7 Jun 2024Qwen2-72BOpen weights3×10²⁴——
4 Feb 2024Qwen1.5-72BOpen weights1.3×10²⁴——
25 Jan 2024Qwen-VL-MaxAPI or hosted access———
30 Nov 2023Qwen-72BOpen weights1.3×10²⁴——
14 Nov 2023Qwen-Audio-ChatOpen weights———
7 Nov 2023mPLUG-Owl2Open weights———
24 Aug 2023Qwen-VLOpen weights———
18 May 2023ONE-PEACEOpen weights1.8×10²⁰——
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). Alibaba's models as CSV.

Alibaba 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.
LensAlibabaOpenAIAnthropicGooglexAIMetaMicrosoftDeepSeekMistral AINvidia
CapabilityEpoch Capabilities Index, its best model · data157.4#5168.8#1168.8#1166.4#3156.6#6161.8#4130.5#10155.9#7150.0#8146.0#9
OutputNotable AI models released in 2025 · chart11#321#17#415#25#54#60#104#61#93#8
MoneyAnnualized revenue, latest report · data—$50B#2Sep 2026$65B#1Jul 2026—$428M#5Sep 2025——$1B#3Sep 2026$1B#3Sep 2026—
ComputeLargest training run with a published estimate, FLOP · chart1.5×10²⁵#7Qwen3-Max, Sep 20259.4×10²⁶#1GPT-6 Astra, Sep 20263.4×10²⁵#5Claude 3.7 Sonnet, Feb 20255×10²⁵#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 20267.7×10²³#10Mixtral 8x7B, Dec 20231.8×10²⁵#6Nemotron-4 340B, Jun 2024
UseUS adults who used its assistant in the past week, Mar 2026 · data—31.0%#13.0%#620.9%#24.5%#57.5%#410.5%#30.6%#7——
Safety & policyLowest hallucination rate of its models (lower is better) · chart89%#889%#826%#250%#322%#187%#7—82%#484%#683%#5
More measures · Epoch AI data
Text Arenabest score, Mar 2026 · data1,449#51,481#41,503#11,494#31,495#21,335#8—1,424#61,416#7—
Valuationlatest closed round · data—$852.0B#2$965.0B#1—$230.0B#3——$50.0B#4$24.0B#5—
Equity raisedin all · data—$182.8B#1$139.4B#2—$37.0B#3——$7.5B#4$6.5B#5—
Stafflatest report · data—4,500#23,000#46,000#1300#63,400#3—300#6350#5—
AI chips ownedH100-equivalents · data———5.47M#10.55M#41.95M#33.32M#2———
Compute availableH100-equivalents, end of 2025 · data—1.74M#11.19M#31.58M#20.63M#51.00M#4————
Weekly use at workemployed US adults, Jul 2026 · data—33.7%#17.7%#524.2%#23.3%#69.8%#414.8%#3———
Model record · Epoch AI estimates
Notable models, 2014–2025Epoch AI's count, as in chart 1.1.727#661#313#7194#17#887#242#47#86#1030#5
Frontier modelsmarked frontier in Epoch AI's database0#815#23#533#13#55#33#50#80#84#4
Open weightsshare of its notable models since 202375%24 of 324%2 of 570%0 of 2214%8 of 5711%1 of 973%19 of 2633%2 of 6100%10 of 1067%4 of 683%15 of 18
Largest estimated training cost2023 US dollars, among models Epoch has costed$109K#10PLUG$366M#2GPT-4.5$26M#5Claude 3.5 Sonnet$31M#4Gemini 1.0 Ultra$388M#1Grok 4$53M#3Llama 3.1-405B$4M#9Megatron-Turing NLG 530B$7M#8DeepSeek-R1 (May 2025)$14M#7Mistral Large$21M#6Nemotron-4 340B
Largest training clusterchips for one model, where Epoch records it512#8CogView100,000#2GPT-6 Astra—57,000#3Gemini 1.0 Ultra200,000#1Grok 432,000#4Llama 4 Behemoth (preview)4,480#6Megatron-Turing NLG 530B2,048#7DeepSeek-V3 (Mar 2025)—6,144#5Nemotron-4 340B
Charts on AI Adoptionthat name the company7332223840414

Every chart that names Alibaba

All charts →
7 charts across 4 hubs name Alibaba or one of its models. Each has its source and a CSV of its data.
AI Benchmarks4 charts
  1. 1,314GPT-5.2 leads the Indic LLM Arena with 1,314, ahead of GPT-5.1 (1,298) and Gemini 3 Flash (1,288)
  2. 22.6%The best of seven AI models read analogue clocks exactly right 22.6% of the time on ClockQA
  3. 50.6%The best AI model read 50.6% of ClockBench clocks correctly in March 2026; humans manage about 90%
  4. 68.3%Kimi K2.5 led CorpFin v2's credit-agreement questions at 68.3% in early 2026; none reached 70%
AI Models1 chart
  1. 21OpenAI released the most notable AI models in 2025: 21, ahead of Google's 15
Robots & Self-Driving Cars1 chart
  1. 0.64GPT-4o scored 0.64 on ResponsibleRobotBench's safe success rate in 2025, the best of five models tested
AI Ethics & Safety1 chart
  1. 1.8%Finix S1 32B hallucinated in 1.8% of document summaries, the lowest of 15 models on Vectara's HHEM test

The other companies, same lens

AI Companies →
OpenAI168.8Epoch Capabilities Index65 ledger entriesOpen tracker →Anthropic168.8Epoch Capabilities Index61 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 →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