Updated 5h ago
AI Leaderboard - 916 models ranked by capability, price & context
Independent benchmark results from Epoch AI and Artificial Analysis, live API prices and context windows from OpenRouter. Each ranking uses one named measure - we don't blend them into a made-up score.
Highest capability
GPT-6 Astra
ECI 166.6Epoch AI
Leads on reasoning
GPT-6 Astra
95.8% GPQAEpoch AI
Wins at coding
GPT-6 Astra
74.1% DeepSWEEpoch AI
Cheapest in the top 10
GPT-5.6 Sol
$4.00/M blendedOpenRouter
Longest context
Grok 4.20
2.0M tokensOpenRouter
Best open weights
Kimi K3
ECI 157.68Epoch AI
Ranked by the Epoch Capabilities Index (ECI) - Epoch AI's statistical model over dozens of benchmarks. Source: Epoch AI. Variants (reasoning effort, batch) are folded into one row.
| # | Model | Capability | Reasoning | Coding | Context | $/M | Weights | Released |
|---|---|---|---|---|---|---|---|---|
| 1 | GPT-6 AstraNEWOpenAI | 166.6 | 95.8 | 73.2 | 1.05M | $20.00 | Closed | 3 Sep 2026 |
| 2 | Claude Fable 5.1NEWAnthropic | 165.0 | — | — | 1M | $20.00 | Closed | 1 Sep 2026 |
| 3 | Claude Fable 5Anthropic | 163.6 | 85.9 | 69.7 | 1M | $20.00 | Closed | 9 Jun 2026 |
| 4 | Claude Opus 5Anthropic | 162.7 | 93.9 | 73.6 | 1M | $10.00 | Closed | 24 Jul 2026 |
| 5 | GPT-5.5 ProOpenAI | 162.4 | — | — | 1.05M | $67.50 | Closed | 23 Apr 2026 |
| 6 | GPT-5.6 SolOpenAI | 162.0 | 93.5 | 72.7 | 1.05M | $4.00 | Closed | 9 Jul 2026 |
| 7 | GPT-5.6 TerraOpenAI | 159.3 | 93.3 | 69.6 | 1.05M | $4.50 | Closed | 9 Jul 2026 |
| 8 | GPT-5.5OpenAI | 159.3 | — | 64.4 | 1.05M | $11.25 | Closed | 23 Apr 2026 |
| 9 | GPT-5.4 ProOpenAI | 159.1 | — | — | 1.05M | $67.50 | Closed | 5 Mar 2026 |
| 10 | Claude Opus 4.8Anthropic | 158.3 | 91.0 | 59.0 | 1M | $10.00 | Closed | 28 May 2026 |
| 11 | Gemini 3.7 FlashGoogle | 157.7 | 94.8 | 65.3 | 1.05M | $1.50 | Closed | 13 Aug 2026 |
| 12 | Kimi K3Moonshot AI | 157.7 | 93.1 | 68.5 | 1.05M | $6.00 | Open | 16 Jul 2026 |
| 13 | Gemini 3.8 FlashNEWGoogle | 157.1 | 95.4 | 73.8 | 1.05M | $1.50 | Closed | 2 Sep 2026 |
| 14 | GPT-5.4OpenAI | 156.9 | 89.9 | — | 1.05M | $5.63 | Closed | 5 Mar 2026 |
| 15 | Muse Spark 1.3NEWMeta | 156.9 | — | — | 1.05M | $2.00 | Closed | 2 Sep 2026 |
— means no published score from that source yet · click a model for every benchmark with its sourceFull leaderboard (268) →
New - awaiting independent evaluation
All new models →- Claude Sonnet 5.5Anthropic · 28 Sep 2026 · AA 56 · 1M · $4.00/M
- Qwen3.8 Max PrimeAlibaba (Qwen) · 23 Sep 2026 · 1M · $6.00/M
- GPT-6 SolOpenAI · 22 Sep 2026 · AA 47.5 · 1.05M · $4.00/M
- GPT-6 LunaOpenAI · 22 Sep 2026 · AA 37.3 · 1.05M · $0.20/M
- GPT-6 Sol ProOpenAI · 22 Sep 2026 · 1.05M · $4.00/M
- GPT-6 Luna ProOpenAI · 22 Sep 2026 · 1.05M · $0.20/M
- Claude Opus 5.5Anthropic · 22 Sep 2026 · AA 57.6 · 1M · $8.00/M
- Grok 4.7xAI · 21 Sep 2026 · AA 46.4 · 500K · $3.00/M
Not ranked until an independent evaluator (Epoch AI) publishes scores. We don't use vendor-reported benchmark claims for rankings.
Best AI for coding →
Models ranked on DeepSWE
Best AI for reasoning →
Ranked on GPQA Diamond
Best AI for math →
Ranked on competition math
Best open-weight models →
Weights you can run yourself
Cheapest capable models →
Lowest blended API price
Best AI coding tools →
Apps & agents by popularity
Best AI image generators →
Apps & models
Best AI for research →
Answer engines & deep research
Find something
Cheap flights, furniture, food nearby, apartments - the right websites in seconds.
Try:
AI products & agents - measured attention, not quality
Most popular AI products · Open source
Signals as of 2026-09-29 · 48 ranked · no traffic estimates usedFull screener →
Popularity rank, 7 days
Biggest movers
Rising
- Qodo47places up
- Imagen & Nano Banana33places up
- Veo & Flow24places up
- Kling AI19places up
- llama.cpp9places up
- Grammarly9places up
Last 7 days vs prior 4 weeks
Trending now
- Scale Data Engine+192%+192% Wikipedia views
- opencode+120%+120% HN stories
- Suno+33%+33% Wikipedia views
- Ray-Ban Meta glasses+27%+27% Wikipedia views
- Vercel AI SDK+26%+26% SDK downloads
- Rabbit r1+25%+25% Wikipedia views
Event engine
Latest AI updates
- Price rise+60%Hy3 API price raised
- Price rise+684%GLM 5.3 API price raised
- Price cut-15%GLM 5.2 API price cut
- Price rise+45%GLM 5.1 API price raised
- Price rise+25%Grok 4.7 API price raised
- Price cut-85%Qwen3.8 27B API price cut
- Price rise+8%Qwen3 30B A3B API price raised
- Price cut-75%gpt-oss-120b API price cut
- Price cut-30%MiniMax M2.7 API price cut
Discovery score
Newly discovered & emerging
- ModelClaude Fable 5.184Anthropic · 1 Sep 2026
- ModelGPT-6 Astra81OpenAI · 3 Sep 2026
- ModelGemini 3.8 Flash76Google · 2 Sep 2026
- ModelDeepSeek V4.1 Flash74DeepSeek · 9 Sep 2026
- ModelGLM 5.3 Flash71Z.ai (Zhipu AI) · 20 Aug 2026
- ModelMuse Spark 1.371Meta · 2 Sep 2026
- ModelGemini 3.7 Flash70Google · 13 Aug 2026
- productScale Data Engine69Scale AI · date unknown
- ModelGLM 5.369Z.ai (Zhipu AI) · 14 Aug 2026
Discovery score = recency, growth, capability and release activity. It is not a popularity measure.
Primary & press sources, deduplicated
AI news
- AnnouncementAnthropic's prospectus details losses, growth, and, yes, a warning that its AI could end humanityTechCrunchAnthropic
- LaunchPeak XV ups Surge seed investment ceiling to $5M, unveils 18-startup cohortTechCrunch
- AnnouncementOpenAI reportedly ditches model over safety concernsTechCrunchOpenAI
- AnnouncementRoundtables: The Deadly Failures of The Virtual Border WallMIT Technology Review
- LaunchMajorGrok 4.7 is now available on Amazon Bedrock
- FundingAMD is acquiring AI company World Labs in a deal worth more than $8 billionThe VergeAMD
- FundingSource: Inference provider Modal Labs closing in on $750M round at $15.75B valuationTechCrunch
- AcquisitionAMD will acquire Fei-Fei Li's World Labs for $8.2 billionTechCrunchAMD
Ecosystem
AI market by category
Code reviewDevelopment
App buildersDevelopment
Voice & speechCreative
AI searchKnowledge
ResearchKnowledge
Notes & knowledgeProductivity
AutomationBusiness
Customer supportBusiness
Sales & marketingBusiness
Model APIsInfrastructure
Local LLM runtimesInfrastructure
Rankings use 3 measured signal types. Every table value links to its source.