# AI Race Scoreboard

> The AI race, scored like a sport. One score per AI lab from model rankings, revenue, valuation and prediction-market odds, with a source behind every number.

AI Race Scoreboard (https://airacescoreboard.com/) ranks the leading AI labs every week with one composite score from 0 to 100. The score is a weighted average of five components: Arena text leaderboard (35), Artificial Analysis Intelligence Index (15), reported revenue run-rate (20), latest closed valuation (10) and Polymarket "best AI model end of 2026" odds (20). Components without reliable public data are left out and the other weights rescaled; the site never fills a gap with its own guess. Third-party estimates are labeled as such on the lab pages.

Current week: 2026-10-17. Leader: Anthropic (94.4). New #1 this week, taking the top spot from Google DeepMind.

## Current standings (week of 2026-10-17)

1. Anthropic: 94.4/100 (5/5 components), up 1. https://airacescoreboard.com/lab/anthropic/
2. Google DeepMind: 93.3/100 (3/5 components), down 1. https://airacescoreboard.com/lab/google/
3. OpenAI: 67.3/100 (5/5 components). https://airacescoreboard.com/lab/openai/
4. Meta Superintelligence Labs: 64.9/100 (3/5 components). https://airacescoreboard.com/lab/meta/
5. Alibaba (Qwen): 50.7/100 (4/5 components). https://airacescoreboard.com/lab/alibaba/
6. xAI (SpaceXAI): 49.4/100 (5/5 components). https://airacescoreboard.com/lab/xai/
7. Moonshot AI (Kimi): 46.3/100 (5/5 components). https://airacescoreboard.com/lab/moonshot/
8. Z.ai (Zhipu): 44.4/100 (3/5 components). https://airacescoreboard.com/lab/zai/
9. DeepSeek: 41.2/100 (5/5 components). https://airacescoreboard.com/lab/deepseek/
10. Mistral AI: 24.6/100 (5/5 components). https://airacescoreboard.com/lab/mistral/

## Methodology

Each lab gets up to five component scores, each 0-100. The overall score is their weighted average over the components that have data:

score = sum(weight_i * component_i) / sum(weight_i)

- arena (weight 35): Arena text score: 100 × (score − (leader − 100)) / 100, clamped to 0–100. The leader gets 100; 100 Elo points behind gets 0.
- aa (weight 15): Artificial Analysis Intelligence Index: 100 × index / best index.
- revenue (weight 20): Revenue run-rate on a log scale: 100 × log10(rev / $100M) / log10(top rev / $100M). $100M = 0.
- valuation (weight 10): Latest closed valuation on a log scale: 100 × log10(val / $1B) / log10(top val / $1B). $1B = 0.
- odds (weight 20): Polymarket "best AI model end of 2026" Yes price: 100 × price / highest price.

Rules:
- Missing data is never guessed. A component with no reliable public number is dropped and the other weights are rescaled. Coverage (for example 3/5) is shown next to every score.
- Known bias: rescaling helps labs inside public giants (Google, Meta), which are scored only on models and odds.
- Claimed vs reported: when a company or its investors push a higher number than what is later reported, both are shown and the reported one is scored. Run-rates use each company's own accounting (Anthropic books some cloud-partner revenue gross, OpenAI books it net). "q x 4" means a quarter annualized by us.
- Compute (GW, capex) and total funding are shown but not scored.
- Weekly arrows compare with the snapshot about 7 days earlier. The comparison snapshot (2026-10-10) is backfilled: real historical Polymarket prices, financials limited to facts dated on or before that day, Arena and AA held at current levels.
- Sponsors never affect scores or rankings.

## Sources

- Arena text leaderboard, Overall (https://arena.ai/leaderboard/text): manual snapshot, as of 2026-10-15. Flagged stale when older than 8 days.
- Artificial Analysis Intelligence Index (https://artificialanalysis.ai/leaderboards/models): fetched 2026-10-17T14:41:44.044Z. "Position" is the rank among non-deprecated models with an index on that page, not an official AA rank.
- Polymarket (public API): fetched 2026-10-17T14:41:41.693Z. Scored market: "Which company has best AI model end of 2026?" (https://polymarket.com/event/which-company-has-best-ai-model-end-of-2026).
- Revenue, valuation, funding, compute: company filings, company posts and major outlets, updated by hand. Every figure below has its basis, as-of date and source link.

## Prediction markets

- Which company has best AI model end of 2026? (https://polymarket.com/event/which-company-has-best-ai-model-end-of-2026), volume $2.56M, resolves 2026-12-31: Anthropic 47.5%, Google 39.5%, OpenAI 7.6%, Meta 2.75%
- Which company has the best AI model end of October? (https://polymarket.com/event/which-company-has-the-best-ai-model-end-of-october), volume $2.69M, resolves 2026-10-31: Google 75.35%, Anthropic 23.5%, OpenAI 0.65%, Meta 0.35%
- Which company has the best AI model end of November? (https://polymarket.com/event/which-company-has-the-best-ai-model-end-of-november), volume $0.30M, resolves 2026-12-01: Anthropic 47.5%, Google 42.5%, OpenAI 6.3%, Meta 2.65%

## Labs in detail

### 1. Anthropic

- Page: https://airacescoreboard.com/lab/anthropic/ | JSON: https://airacescoreboard.com/data/v1/labs/anthropic.json
- HQ: San Francisco, US
- Score: 94.4/100, coverage 5/5, rank change 1, score change 0.5
- Components (0-100): arena 84, aa 100, revenue 100, valuation 100, odds 100
- Arena text: #3, Claude Opus 5.5 (High) (1508)
- Artificial Analysis Intelligence Index: 57.6223698102963, Claude Opus 5.5 (Max, Default Fallback) (position 1)
- Polymarket 2026 #1 odds: 47.5% (+6pp on the week)
- Revenue (reported): $65B, annualized run-rate (gross accounting), as of 2026-07-31. Source: Reuters https://www.reuters.com/technology/anthropic-revenue-run-rate-tops-65-billion-source-says-2026-08-17/. Note: Run-rate 'topped $65B by end of July', per a source. Anthropic books cloud-partner sales gross, so it is not directly comparable to OpenAI's figure.
- Valuation: $965B, post-money, Series H, as of 2026-05-28. Source: Anthropic https://www.anthropic.com/news/series-h
- Total funding (not scored): $133B, as of 2026-10-17. Source: StockAnalysis (funding database) https://stockanalysis.com/private/anthropic/valuation/
- Compute (not scored): Compute contracted since Oct 2025: ≥14.8 GW, as of 2026-09-06. Source: The Information via AI-RTZ https://michaelparekh.substack.com/p/ai-anthropics-517b-vs-openais-750b
- Compute (not scored): Anthropic-Amazon New Carlisle, current IT power (est.): ~910 MW, as of 2026-10-15. Source: Epoch AI https://epoch.ai/graphs/largest-ai-data-centers-by-power-capacity
- Timeline:
  - 2026-10-14: Q3 2026 revenue $16.8B, up 45% from $11.6B in Q2. Profitable on an adjusted basis. (https://www.cnbc.com/2026/10/15/anthropics-third-quarter-revenue-hits-16point8-billion-ipo.html)
  - 2026-10-14: 11% adjusted margin; about $1.70 of revenue per $1 spent on compute. (https://www.ft.com/content/11632d03-358c-454b-af95-d06d4072932b)
  - 2026-09-06: Compute deals worth up to $517B signed in 11 months (The Information tally). (https://michaelparekh.substack.com/p/ai-anthropics-517b-vs-openais-750b)

### 2. Google DeepMind

- Page: https://airacescoreboard.com/lab/google/ | JSON: https://airacescoreboard.com/data/v1/labs/google.json
- HQ: London / Mountain View | Parent: Alphabet (public)
- Score: 93.3/100, coverage 3/5, rank change -1, score change -4.8
- Components (0-100): arena 100, aa 91.2, revenue n/a, valuation n/a, odds 83.2
- Arena text: #1, Gemini 4 Argon (High) (1524)
- Artificial Analysis Intelligence Index: 52.5605655745982, Gemini 4 Argon (High) (position 9)
- Polymarket 2026 #1 odds: 39.5% (-3pp on the week)
- Revenue (reported): no reliable public figure
- Valuation: no reliable public figure
- Total funding (not scored): no reliable public figure
- Compute (not scored): Alphabet 2026 capex guidance (company-wide): $195–205B, as of 2026-07-22. Source: CNBC (Q2 2026 earnings) https://www.cnbc.com/2026/07/22/google-earnings-q2-goog-live-updates.html
- Timeline:
  - 2026-07-22: Gemini app reached 950M monthly active users. Google Cloud Q2 revenue $24.8B (+82% YoY). (https://www.sec.gov/Archives/edgar/data/1652044/000165204426000066/googexhibit991q22026.htm)
- Notes: This is a division of a public company. Alphabet does not disclose lab-level revenue or valuation, so those score components are excluded and the remaining weights are rescaled.

### 3. OpenAI

- Page: https://airacescoreboard.com/lab/openai/ | JSON: https://airacescoreboard.com/data/v1/labs/openai.json
- HQ: San Francisco, US
- Score: 67.3/100, coverage 5/5, rank change 0, score change -1.5
- Components (0-100): arena 61, aa 91.4, revenue 95.9, valuation 98.2, odds 16
- Arena text: #21, GPT-5.6 Sol (xHigh) (1485)
- Artificial Analysis Intelligence Index: 52.673669395513, GPT-6 Astra (Max) (position 8)
- Polymarket 2026 #1 odds: 7.6% (-2.4pp on the week)
- Revenue (reported): $50B, annualized run-rate, end of September (net of partner share), as of 2026-09-30. Source: Axios https://www.axios.com/2026/10/08/openai-50-billion-arr-anthropic-revenue-accounting. Note: OpenAI books only its retained share of some partner (Azure) sales.
- Revenue (claimed, not scored): $70B, earlier investor-facing 'grossed-up' annualized figure, as of 2026-10-08. Source: Axios https://www.axios.com/2026/10/08/openai-50-billion-arr-anthropic-revenue-accounting. Note: Axios: annualized revenue is ~$20B less than the figure previously reported. OpenAI also told investors it expects ≥$70B annualized by end of 2026 (Bloomberg via Yahoo Finance).
- Valuation: $852B, post-money, $122B round, as of 2026-03-31. Source: OpenAI https://openai.com/index/accelerating-the-next-phase-ai/
- Valuation (in talks, not scored): $1.4T, pre-money, in talks to raise ≥$30B (not closed), as of 2026-10-12. Source: TradedVC / Bloomberg https://traded.co/vc/deal/openai-seeks-to-raise-30-billion-at-1-4-trillion-valuation-with-investor-support/
- Total funding (not scored): $178B, as of 2026-10-16. Source: StockAnalysis (funding database) https://stockanalysis.com/private/openai/valuation/. Note: Other databases differ: VC Market Cap lists $192.20B.
- Compute (not scored): Planned capacity by 2030 (told investors): 30 GW, as of 2026-09-09. Source: AI-RTZ https://michaelparekh.substack.com/p/ai-anthropics-517b-vs-openais-750b
- Compute (not scored): Compute spend forecast through 2030: ~$750B, as of 2026-09-09. Source: AI-RTZ (The Information) https://michaelparekh.substack.com/p/ai-anthropics-517b-vs-openais-750b
- Timeline:
  - 2026-10-12: In talks to raise ≥$30B at a $1.4T pre-money valuation (not closed). (https://traded.co/vc/deal/openai-seeks-to-raise-30-billion-at-1-4-trillion-valuation-with-investor-support/)
  - 2026-10-08: The ~$70B figure became ~$50B. Annualized revenue is about $20B less than previously reported (an accounting 'gross-up'). (https://www.axios.com/2026/10/08/openai-50-billion-arr-anthropic-revenue-accounting)
  - 2026-09-09: Expects to burn about $25B of cash in 2026 (The Information). (https://michaelparekh.substack.com/p/ai-anthropics-517b-vs-openais-750b)

### 4. Meta Superintelligence Labs

- Page: https://airacescoreboard.com/lab/meta/ | JSON: https://airacescoreboard.com/data/v1/labs/meta.json
- HQ: Menlo Park, US | Parent: Meta Platforms (public)
- Score: 64.9/100, coverage 3/5, rank change 0, score change 0.2
- Components (0-100): arena 85, aa 96.8, revenue n/a, valuation n/a, odds 5.8
- Arena text: #2, Muse Wonder (Max) (1509)
- Artificial Analysis Intelligence Index: 55.8021199855616, Muse Wonder Full (Max) (position 4)
- Polymarket 2026 #1 odds: 2.75% (+0.6pp on the week)
- Revenue (reported): no reliable public figure
- Valuation: no reliable public figure
- Total funding (not scored): no reliable public figure
- Compute (not scored): Meta 2026 capex guidance (company-wide): $130–145B, as of 2026-07-29. Source: Meta Q2 2026 results https://investor.atmeta.com/investor-news/press-release-details/2026/Meta-Reports-Second-Quarter-2026-Results/default.aspx
- Compute (not scored): Meta Prometheus, current IT power (est.): ~471 MW, as of 2026-10-15. Source: Epoch AI https://epoch.ai/graphs/largest-ai-data-centers-by-power-capacity
- Timeline:
  - 2026-07-29: 2026 capex guidance is $130–145B. (https://investor.atmeta.com/investor-news/press-release-details/2026/Meta-Reports-Second-Quarter-2026-Results/default.aspx)
- Notes: This is a division of a public company. Lab-level revenue and valuation are not disclosed.

### 5. Alibaba (Qwen)

- Page: https://airacescoreboard.com/lab/alibaba/ | JSON: https://airacescoreboard.com/data/v1/labs/alibaba.json
- HQ: Hangzhou, CN | Parent: Alibaba Group (public)
- Score: 50.7/100, coverage 4/5, rank change 0, score change 0
- Components (0-100): arena 58, aa 78.8, revenue 66.2, valuation n/a, odds 1.2
- Arena text: #24, Qwen3.8-Max (1482)
- Artificial Analysis Intelligence Index: 45.4152084980521, Qwen3.8 Max (0902) (position 29)
- Polymarket 2026 #1 odds: 0.55%
- Revenue (reported): $7.3B, AI-related product revenue, June quarter × 4 (computed; parent-segment metric), as of 2026-06-30. Source: Alibaba June Quarter 2026 results https://data.alibabagroup.com/ecms-files/1532295521/fa5d65fc-9b3e-4e82-a8fc-4ce1c3e2c407/Alibaba%20Group%20Announces%20June%20Quarter%202026%20Results.pdf. Note: Quarter: RMB12.376B (US$1.824B). This is Alibaba's AI-related product revenue line, not Qwen alone.
- Valuation: no reliable public figure
- Total funding (not scored): no reliable public figure
- Compute (not scored): Group capex, June quarter: US$9.975B, as of 2026-06-30. Source: Alibaba June Quarter 2026 results https://data.alibabagroup.com/ecms-files/1532295521/fa5d65fc-9b3e-4e82-a8fc-4ce1c3e2c407/Alibaba%20Group%20Announces%20June%20Quarter%202026%20Results.pdf
- Timeline:
  - 2026-06-30: AI-related product revenue: 12th straight quarter of triple-digit growth. (https://data.alibabagroup.com/ecms-files/1532295521/fa5d65fc-9b3e-4e82-a8fc-4ce1c3e2c407/Alibaba%20Group%20Announces%20June%20Quarter%202026%20Results.pdf)
  - 2026-06-30: AI Cloud & Compute Services revenue US$7.139B in the quarter (+45% YoY). (https://data.alibabagroup.com/ecms-files/1532295521/fa5d65fc-9b3e-4e82-a8fc-4ce1c3e2c407/Alibaba%20Group%20Announces%20June%20Quarter%202026%20Results.pdf)
- Notes: Alibaba is a public parent. Valuation and funding are not applicable at the lab level.

### 6. xAI (SpaceXAI)

- Page: https://airacescoreboard.com/lab/xai/ | JSON: https://airacescoreboard.com/data/v1/labs/xai.json
- HQ: Palo Alto / Memphis, US | Parent: SpaceX (acquired Feb 2026)
- Score: 49.4/100, coverage 5/5, rank change 0, score change -0.3
- Components (0-100): arena 51, aa 80.6, revenue 53.9, valuation 80.3, odds 3.5
- Arena text: #37, Grok 4.20 Beta 1 (1475)
- Artificial Analysis Intelligence Index: 46.4465506302286, Grok 4.7 (Xhigh) (position 25)
- Polymarket 2026 #1 odds: 1.65% (-0.4pp on the week)
- Revenue (reported): $3.3B, Q1 2026 segment revenue × 4 (computed; includes X ads/subscriptions), as of 2026-03-31. Source: Morningstar (SpaceX S-1) https://www.morningstar.com/stocks/financials-look-reckless-lifting-xais-hood-spacex-ipo. Note: Q1 2026 revenue $818M, operating loss $2.47B. We multiply the quarter by 4 ourselves.
- Valuation: $250B, implied value in the SpaceX all-stock merger, as of 2026-02-02. Source: Reuters https://www.reuters.com/business/musks-spacex-merge-with-xai-combined-valuation-125-trillion-bloomberg-news-2026-02-02/
- Total funding (not scored): no reliable public figure
- Compute (not scored): Colossus 2, current IT power (est.), largest tracked: ~946 MW, as of 2026-10-15. Source: Epoch AI https://epoch.ai/graphs/largest-ai-data-centers-by-power-capacity
- Compute (not scored): AI-segment capex, Q1 2026: $7.7B, as of 2026-03-31. Source: TechCrunch (S-1) https://techcrunch.com/2026/05/20/xai-burned-6-4b-last-year-spacexs-ipo-filing-shows-why-the-spending-is-far-from-over/
- Timeline:
  - 2026-05-20: Lost $6.4B from operations on $3.2B revenue in 2025 (SpaceX S-1). (https://techcrunch.com/2026/05/20/xai-burned-6-4b-last-year-spacexs-ipo-filing-shows-why-the-spending-is-far-from-over/)
  - 2026-05-20: Grok: 117M monthly users, but only 1.9M pay for the advanced models. (https://www.morningstar.com/stocks/financials-look-reckless-lifting-xais-hood-spacex-ipo)

### 7. Moonshot AI (Kimi)

- Page: https://airacescoreboard.com/lab/moonshot/ | JSON: https://airacescoreboard.com/data/v1/labs/moonshot.json
- HQ: Beijing, CN
- Score: 46.3/100, coverage 5/5, rank change 0, score change -0.2
- Components (0-100): arena 63, aa 75.7, revenue 35.6, valuation 56.9, odds 0.7
- Arena text: #18, Kimi K3 (Max) (1487)
- Artificial Analysis Intelligence Index: 43.5938229518782, Kimi K3 (Max) (position 33)
- Polymarket 2026 #1 odds: 0.35% (-0.2pp on the week)
- Revenue (reported): $1B, ARR (current, per Bloomberg), as of 2026-10-06. Source: Silicon Republic (Bloomberg) https://www.siliconrepublic.com/start-ups/bloomberg-moonshot-closes-round-at-50bn-valuation-before-2027-ipo
- Revenue (claimed, not scored): $2B, ARR expected by December, as of 2026-10-06. Source: Silicon Republic (Bloomberg) https://www.siliconrepublic.com/start-ups/bloomberg-moonshot-closes-round-at-50bn-valuation-before-2027-ipo. Note: This is a forward-looking expectation, not a reported number.
- Valuation: $50B, final private round (reported), as of 2026-10-06. Source: Silicon Republic (Bloomberg) https://www.siliconrepublic.com/start-ups/bloomberg-moonshot-closes-round-at-50bn-valuation-before-2027-ipo
- Total funding (not scored): no reliable public figure

- Timeline:
  - 2026-10-06: Closed its final private round at a $50B valuation; Hong Kong IPO planned for early 2027. ARR $1B, targeting $2B by December. (https://www.siliconrepublic.com/start-ups/bloomberg-moonshot-closes-round-at-50bn-valuation-before-2027-ipo)

### 8. Z.ai (Zhipu)

- Page: https://airacescoreboard.com/lab/zai/ | JSON: https://airacescoreboard.com/data/v1/labs/zai.json
- HQ: Beijing, CN
- Score: 44.4/100, coverage 3/5, rank change 0, score change 0
- Components (0-100): arena 55, aa 77.7, revenue n/a, valuation n/a, odds 0.7
- Arena text: #28, GLM-5.3 (Max) (1479)
- Artificial Analysis Intelligence Index: 44.777392385614, GLM-5.3 (Max) (position 31)
- Polymarket 2026 #1 odds: 0.35%
- Revenue (reported): no reliable public figure
- Valuation: no reliable public figure
- Total funding (not scored): no reliable public figure

- Notes: No sourced financials gathered yet. Only model and odds components are scored.

### 9. DeepSeek

- Page: https://airacescoreboard.com/lab/deepseek/ | JSON: https://airacescoreboard.com/data/v1/labs/deepseek.json
- HQ: Hangzhou, CN
- Score: 41.2/100, coverage 5/5, rank change 0, score change 0
- Components (0-100): arena 51, aa 68.5, revenue 35.6, valuation 57.4, odds 1.1
- Arena text: #38, DeepSeek-V4.1-Flash (Max) (1475)
- Artificial Analysis Intelligence Index: 39.456167472527, DeepSeek V4.1 Flash (Max) (position 47)
- Polymarket 2026 #1 odds: 0.5% (+0.1pp on the week)
- Revenue (reported): $1B, annualized revenue, September (third-party estimate), as of 2026-09-30. Source: Sacra (estimate) https://sacra.com/c/deepseek/. Note: This is an estimate, not a company disclosure.
- Valuation: $51.8B, post-money of first external round (~RMB350.88B), as of 2026-07-17. Source: Caixin / Sacra https://www.caixinglobal.com/2026-07-17/deepseek-reaches-52-billion-valuation-in-round-backed-by-tencent-catl-102465358.html
- Valuation (in talks, not scored): no reliable public figure
- Total funding (not scored): $7.4B, as of 2026-07-17. Source: Caixin https://www.caixinglobal.com/2026-07-17/deepseek-reaches-52-billion-valuation-in-round-backed-by-tencent-catl-102465358.html. Note: This is the first external round. Earlier funding from High-Flyer is not included.

- Timeline:
  - 2026-10-06: Weighing a second raise of up to RMB100B at ~RMB500B valuation. (https://www.cnbc.com/2026/10/06/deepseek-funding-round.html)
  - 2026-09-30: Sacra estimates about $1B annualized revenue in September 2026 (not a company disclosure). (https://sacra.com/c/deepseek/)

### 10. Mistral AI

- Page: https://airacescoreboard.com/lab/mistral/ | JSON: https://airacescoreboard.com/data/v1/labs/mistral.json
- HQ: Paris, FR
- Score: 24.6/100, coverage 5/5, rank change 0, score change 0
- Components (0-100): arena 8, aa 66.6, revenue 35.6, valuation 46.2, odds 0.5
- Arena text: #111, Mistral Large 4 (1432)
- Artificial Analysis Intelligence Index: 38.3770408917494, Mistral Large 4 Preview (position 49)
- Polymarket 2026 #1 odds: 0.25%
- Revenue (reported): $1B, ARR, 'slightly above $1B' per CFO, as of 2026-09-08. Source: WSJ https://www.wsj.com/tech/ai/mistral-ai-exceeds-24-billion-valuation-after-samsung-led-investment-round-ea377b82
- Valuation: $24B, post-money, ~€21B, Series D, as of 2026-09-08. Source: Reuters https://www.reuters.com/world/europe/french-ai-company-mistral-hits-24-billion-valuation-funding-round-2026-09-08/
- Total funding (not scored): $7.5B, as of 2026-09-08. Source: Crunchbase News https://news.crunchbase.com/venture/europe-record-setting-mistral-ai-raise/

- Timeline:
  - 2026-09-08: Raised €3B (~$3.5B) led by Samsung at ~$24B. The CFO says ARR is slightly above $1B. (https://www.wsj.com/tech/ai/mistral-ai-exceeds-24-billion-valuation-after-samsung-led-investment-round-ea377b82)

## Data (JSON, CORS open, no key)

- [Standings v1](https://airacescoreboard.com/data/v1/standings.json): this week's ranks, scores, components, weekly changes, weights, formula
- [Lab facts v1](https://airacescoreboard.com/data/v1/labs.json): every sourced fact per lab, with basis, as_of and source_url
- Per-lab JSON: https://airacescoreboard.com/data/v1/labs/{id}.json (ids: anthropic, google, openai, meta, alibaba, xai, moonshot, zai, deepseek, mistral)
- [History index v1](https://airacescoreboard.com/data/v1/history/index.json): weekly snapshots
- [API docs](https://airacescoreboard.com/data/): fields, versioning, how to cite
- [RSS feed](https://airacescoreboard.com/feed.xml)

## Pages

- [Standings](https://airacescoreboard.com/)
- [Methodology](https://airacescoreboard.com/methodology/): formula, weights, sources, known bias, disclaimer
- [Anthropic](https://airacescoreboard.com/lab/anthropic/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Google DeepMind](https://airacescoreboard.com/lab/google/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [OpenAI](https://airacescoreboard.com/lab/openai/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Meta Superintelligence Labs](https://airacescoreboard.com/lab/meta/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Alibaba (Qwen)](https://airacescoreboard.com/lab/alibaba/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [xAI (SpaceXAI)](https://airacescoreboard.com/lab/xai/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Moonshot AI (Kimi)](https://airacescoreboard.com/lab/moonshot/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Z.ai (Zhipu)](https://airacescoreboard.com/lab/zai/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [DeepSeek](https://airacescoreboard.com/lab/deepseek/): score breakdown, model ranks, sourced revenue, valuation, funding and compute
- [Mistral AI](https://airacescoreboard.com/lab/mistral/): score breakdown, model ranks, sourced revenue, valuation, funding and compute

## Update cadence

Weekly, on Mondays (Europe/Bucharest). Polymarket and Artificial Analysis are fetched on each run; Arena is a manual snapshot; revenue, valuation and funding are updated by hand when a sourced figure changes.

## License

Data is licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may reuse and adapt it, also commercially, if you credit AI Race Scoreboard with a link to https://airacescoreboard.com. Every v1 JSON file has a license field.

## How to cite

AI Race Scoreboard, weekly AI lab standings, week of 2026-10-17. https://airacescoreboard.com/

## Optional

- [Full text for LLMs](https://airacescoreboard.com/llms-full.txt): methodology, standings and every sourced fact in one file
- [Sponsor](https://airacescoreboard.com/sponsor/): sponsors never affect scores

## Disclaimer

Not affiliated with any AI lab, Arena, Artificial Analysis or Polymarket. Not investment, trading or betting advice. Data can be wrong or stale; every figure links to its source. Prediction-market prices are probabilities implied by traders, not facts.