Release watch
What arrived, what got measured, and who is due. Rebuilt every morning from the day’s data.
Data as of 9 Sept 2026
New in the last two weeks
| Model | Released | Index | Rank | Price $/M | Status |
|---|---|---|---|---|---|
| KAT-Coder-Pro V2KwaiKAT | seen 9 Sept 2026 | 58.1 | #131 | $0.525 | Ranked |
 Grok 3 Mini FasthighxAI | seen 9 Sept 2026 | 53.5 | #222 | – | Ranked |
| JT-35B-FlashChina Mobile | seen 9 Sept 2026 | 52.1 | #259 | – | Ranked |
 Nemotron Cascade 2 30B A3BNVIDIA | seen 9 Sept 2026 | 51.0 | #275 | – | Ranked |
| K-EXAONELG AI Research | seen 9 Sept 2026 | 50.9 | #276 | – | Ranked |
| Grok 3 Mini FastlowxAI | seen 9 Sept 2026 | 50.3 | #288 | – | Ranked |
| EXAONE 4.5 33BLG AI Research | seen 9 Sept 2026 | 50.0 | #298 | – | Ranked |
| K2 Think V2thinkingMBZUAI Institute of Foundation Models | seen 9 Sept 2026 | 49.3 | #309 | – | Ranked |
 Gemini 2.5 Flash Lite 09thinkingGoogle | seen 9 Sept 2026 | 48.8 | #314 | $0.175 | Ranked |
 Nova 2.0 Omni (medium)mediumAmazon | seen 9 Sept 2026 | 48.5 | #323 | $0.850 | Ranked |
| Ring 1tUnknown | seen 9 Sept 2026 | 48.0 | #340 | – | Ranked |
| DiffusionGemma 26B A4BGoogle | seen 9 Sept 2026 | 46.5 | #369 | – | Ranked |
 Solar Open 100BthinkingUpstage | seen 9 Sept 2026 | 46.2 | #371 | – | Ranked |
| Ling 1tUnknown | seen 9 Sept 2026 | 45.1 | #399 | – | Ranked |
| MiniCPM5-1BOpenBMB | seen 9 Sept 2026 | 44.9 | #403 | – | Ranked |
 LFM2.5-8B-A1BLiquid AI | seen 9 Sept 2026 | 44.3 | #419 | – | Ranked |
| Nemotron 3 Nano 4BNVIDIA | seen 9 Sept 2026 | 44.1 | #425 | – | Ranked |
| Tri-21B-ThinkthinkingTrillion Labs | seen 9 Sept 2026 | 44.0 | #427 | – | Ranked |
| LongCat Flash LiteLongCat | seen 9 Sept 2026 | 43.9 | #431 | – | Ranked |
| Motif-2-12.7BMotif Technologies | seen 9 Sept 2026 | 43.9 | #432 | – | Ranked |
| HyperCLOVA X SEED Think (32B)thinkingNaver | seen 9 Sept 2026 | 43.6 | #434 | – | Ranked |
| JT-MINIChina Mobile | seen 9 Sept 2026 | 43.6 | #435 | – | Ranked |
| Falcon-H1R-7BTII UAE | seen 9 Sept 2026 | 43.1 | #444 | – | Ranked |
 Step3 VL 10BStepFun | seen 9 Sept 2026 | 43.0 | #447 | – | Ranked |
| Mi:dm K 2.5 ProKorea Telecom | seen 9 Sept 2026 | 43.0 | #448 | – | Ranked |
 Jamba Reasoning 3BthinkingAI21 Labs | seen 9 Sept 2026 | 42.3 | #463 | – | Ranked |
| Gemma 4 E2BGoogle | seen 9 Sept 2026 | 42.2 | #465 | – | Ranked |
| Jamba 1.7 LargeAI21 Labs | seen 9 Sept 2026 | 41.6 | #483 | – | Ranked |
| Hermes 4 70BthinkingNous Research | seen 9 Sept 2026 | 41.4 | #486 | – | Ranked |
| Nanbeige4.1-3BNanbeige | seen 9 Sept 2026 | 41.3 | #490 | – | Ranked |
| Sarvam 105B (high)Sarvam | seen 9 Sept 2026 | 40.4 | #509 | $0.073 | Ranked |
| LFM2.5-1.2B-InstructLiquid AI | seen 9 Sept 2026 | 39.9 | #517 | – | Ranked |
| Olmo 3 7B ThinkthinkingAllen Institute for AI | seen 9 Sept 2026 | 39.8 | #520 | $0.125 | Ranked |
 Qwen3.5 2BAlibaba | seen 9 Sept 2026 | 39.7 | #525 | – | Ranked |
| Jamba 1.7 MiniAI21 Labs | seen 9 Sept 2026 | 39.5 | #527 | – | Ranked |
 Granite 4.1 3BIBM | seen 9 Sept 2026 | 39.3 | #532 | – | Ranked |
| Gemma 3 270MGoogle | seen 9 Sept 2026 | 39.0 | #538 | – | Ranked |
| Reka Flash 3Reka AI | seen 9 Sept 2026 | 38.9 | #541 | $0.350 | Ranked |
 Llama 4 Maverick BasicMeta | seen 9 Sept 2026 | 38.8 | #542 | – | Ranked |
| Llama Llama 3.3 70B TurboMeta | seen 9 Sept 2026 | 38.6 | #545 | – | Ranked |
| Apertus 70B InstructSwiss AI Initiative | seen 9 Sept 2026 | 38.6 | #546 | $1.34 | Ranked |
| Granite 4.0 H 1BIBM | seen 9 Sept 2026 | 38.5 | #547 | – | Ranked |
| Sarvam 30B (high)Sarvam | seen 9 Sept 2026 | 38.5 | #549 | $0.050 | Ranked |
| LFM2 2.6BLiquid AI | seen 9 Sept 2026 | 38.4 | #550 | – | Ranked |
| Exaone 4.0 1.2BLG AI Research | seen 9 Sept 2026 | 37.3 | #563 | – | Ranked |
| Apertus 8B InstructSwiss AI Initiative | seen 9 Sept 2026 | 37.2 | #565 | $0.125 | Ranked |
| Granite 4.0 1BIBM | seen 9 Sept 2026 | 37.2 | #567 | – | Ranked |
 Tiny Aya GlobalCohere | seen 9 Sept 2026 | 36.6 | #576 | – | Ranked |
| Granite 4.0 H 350MIBM | seen 9 Sept 2026 | 36.5 | #577 | – | Ranked |
| Molmo2-8BAllen Institute for AI | seen 9 Sept 2026 | 36.3 | #583 | – | Ranked |
| MiniCPM-V 4.6 1.3BOpenBMB | seen 9 Sept 2026 | 36.1 | #584 | – | Ranked |
| Qwen3.5 0.8BAlibaba | seen 9 Sept 2026 | 35.9 | #588 | – | Ranked |
| Llama Llama 4 Scout 17B 16eMeta | seen 9 Sept 2026 | 35.8 | #590 | – | Ranked |
| LFM2.5-VL-1.6BLiquid AI | seen 9 Sept 2026 | 34.9 | #599 | – | Ranked |
| Jamba Large 1.6AI21 Labs | seen 9 Sept 2026 | 25.2 | #609 | $3.50 | Ranked |
| Jamba Mini 1.6AI21 Labs | seen 9 Sept 2026 | 21.5 | #610 | $0.250 | Ranked |
| Agnes 2.5 Pro BetaSapiens AI | seen 9 Sept 2026 | – | – | $0.150 | First results arriving |
| Apodex 1.1Apodex | seen 9 Sept 2026 | – | – | $0.975 | First results arriving |
| Apriel-v1.6-15B-ThinkerServiceNow | seen 9 Sept 2026 | – | – | – | First results arriving |
| Cogito v2.1thinkingDeep Cogito | seen 9 Sept 2026 | – | – | $1.25 | First results arriving |
| EXAONE 4.0 32BthinkingLG AI Research | seen 9 Sept 2026 | – | – | – | First results arriving |
 GPT 5.1 InstantOpenAI | seen 9 Sept 2026 | – | – | – | First results arriving |
| JT-4.1 Flash 236B A21BChina Mobile | seen 9 Sept 2026 | – | – | – | First results arriving |
| Kimi Linear 48B A3B InstructKimi | seen 9 Sept 2026 | – | – | – | First results arriving |
| Langston Nim Nvidia Llama 3.3 Nemotron Super 49B v1Unknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Langston Nim Nvidia Llama 3.3 Nemotron Super 49B v1 42e84561thinkingUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| LFM2 8B A1BLiquid AI | seen 9 Sept 2026 | – | – | – | First results arriving |
| A.X-K2SK Telecom | seen 9 Sept 2026 | – | – | – | First results arriving |
| G9v3-39A5BAI9Stars | seen 9 Sept 2026 | – | – | – | First results arriving |
| G9v3-3BAI9Stars | seen 9 Sept 2026 | – | – | – | First results arriving |
| K-EXAONE 2.0LG AI Research | seen 9 Sept 2026 | – | – | – | First results arriving |
| K2 Horizon 375B A23BMBZUAI Institute of Foundation Models | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama 2 70B SteerlmNVIDIA | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama Meta Llama 3.1 70B TurboMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama Meta Llama 3.1 8B TurboMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| MiniCPM5-2BOpenBMB | seen 9 Sept 2026 | – | – | – | First results arriving |
| Motif 3 (Beta)Motif Technologies | seen 9 Sept 2026 | – | – | – | First results arriving |
| Qed NanoLM-Provers | seen 9 Sept 2026 | – | – | – | First results arriving |
| Solar Open2 250BUpstage | seen 9 Sept 2026 | – | – | – | First results arriving |
| Zephyr Orpo 141B A35bHugging Face | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama Meta Llama 3.1 405B TurboMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| Nemotron Lightning 3.5 30B A3bNVIDIA | seen 9 Sept 2026 | – | – | – | First results arriving |
| Qwen Qwen2 5 72B TurboAlibaba | seen 9 Sept 2026 | – | – | – | First results arriving |
| Afm 4 5BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Anubis 70BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
 Claude 37 SonnetthinkingAnthropic | seen 9 Sept 2026 | – | – | – | First results arriving |
| Claude Haiku 3Anthropic | seen 9 Sept 2026 | – | – | – | First results arriving |
| Claude Haiku 4thinkingAnthropic | seen 9 Sept 2026 | – | – | – | First results arriving |
| Claude Opus 4.1 AnthropicAnthropic | seen 9 Sept 2026 | – | – | – | First results arriving |
| Claude Opus 4.8 Claude CodeAnthropic | seen 9 Sept 2026 | – | – | – | First results arriving |
| Cogito v2 Deepseek 671BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Cogito v2 Llama 4 ScoutUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Cogito v2 Llama 405BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| DeepHermes 3 - Llama-3.1 8BNous Research | seen 9 Sept 2026 | – | – | – | First results arriving |
| DeepHermes 3 - Mistral 24BNous Research | seen 9 Sept 2026 | – | – | – | First results arriving |
| Gemini 1.0 Pro 002Google | seen 9 Sept 2026 | – | – | – | First results arriving |
| Gemini 2.5 Deep ThinkthinkingGoogle | seen 9 Sept 2026 | – | – | – | First results arriving |
| Gemini Flash 2.0Google | seen 9 Sept 2026 | – | – | – | First results arriving |
| Gemma 3 27B It FastGoogle | seen 9 Sept 2026 | – | – | – | First results arriving |
 Glm 4 32BZhipu AI | seen 9 Sept 2026 | – | – | $0.100 | First results arriving |
| GPT 5 AgenthighOpenAI | seen 9 Sept 2026 | – | – | – | First results arriving |
| GPT 5.5 CodexOpenAI | seen 9 Sept 2026 | – | – | – | First results arriving |
| GPT 5.5 FactoryOpenAI | seen 9 Sept 2026 | – | – | – | First results arriving |
| Grok 4 Fast RxAI | seen 9 Sept 2026 | – | – | – | First results arriving |
| Jamba Large 1.7AI21 Labs | seen 9 Sept 2026 | – | – | $3.50 | First results arriving |
| Jamba Mini 1.7AI21 Labs | seen 9 Sept 2026 | – | – | $0.250 | First results arriving |
| Llama 3 405BMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama 4 Maverick 17BMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama Llama 2 70BMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
 Mercury CoderInception | seen 9 Sept 2026 | – | – | – | First results arriving |
 Minimax 2.1MiniMax | seen 9 Sept 2026 | – | – | – | First results arriving |
| Openhands Lm 32BAll Hands AI | seen 9 Sept 2026 | – | – | – | First results arriving |
| Qwen Qwen2 5 7B TurboAlibaba | seen 9 Sept 2026 | – | – | – | First results arriving |
| Qwen3 32B FastthinkingAlibaba | seen 9 Sept 2026 | – | – | – | First results arriving |
| Qwen3 5 397BAlibaba | seen 9 Sept 2026 | – | – | – | First results arriving |
| Togethercomputer Llama 2 13BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Togethercomputer Llama 2 7BUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
 Ui Tars 1.5 7BByteDance | seen 9 Sept 2026 | – | – | – | First results arriving |
| Valkyrie 49B v1Unknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Virtuoso LargeUnknown | seen 9 Sept 2026 | – | – | – | First results arriving |
| Llama3-SWE-RL-70BMeta | seen 9 Sept 2026 | – | – | – | First results arriving |
| GPT-6 AstramaxOpenAI | 3 Sept 2026 | 70.6 | #1 | $22.00 | Ranked |
| Muse Spark 1.3xhighMeta | 2 Sept 2026 | 63.9 | #41 | $2.00 | Ranked |
| Gemini 3.8 FlashmediumGoogle | 2 Sept 2026 | 63.5 | #45 | $1.50 | Ranked |
| GPT 6 Astra Provider Adapterno reasoningOpenAI | 2 Sept 2026 | – | – | – | Released; awaiting independent results |
| Muse Spark 1.3 ContributorMeta | 2 Sept 2026 | – | – | $0.125 | Released; awaiting independent results |
| Claude Fable 5.1xhighAnthropic | 1 Sept 2026 | 70.3 | #3 | $20.00 | Ranked |
| Qwen3.8-Flash-NextAlibaba | 26 Aug 2026 | 59.3 | #108 | $0.230 | Ranked |
| Qwen3.8 FlashAlibaba | 26 Aug 2026 | – | – | $0.189 | Released; awaiting independent results |
| Quasar 438B (max, based on GLM-5.2)Multiverse Computing | 10 Aug 2026 | – | – | $0.900 | First results arriving |
| Ling 3.0 TinyInclusionAI | 6 Aug 2026 | – | – | – | First results arriving |
| LFM2.5-2.6BLiquid AI | 4 Aug 2026 | – | – | – | First results arriving |
| Agnes 2.5 Pro AlphaSapiens AI | 24 Jul 2026 | – | – | $0.563 | First results arriving |
| Celeris-1Celeris | 24 Jul 2026 | – | – | $0.325 | First results arriving |
First results this week
Benchmarks that published a first score for a model in the last seven days.
- GPT-5.6 Sol
- AA Intelligence Index: 42.59 Sept 2026
- GPQA Diamond (AA): 92.8%9 Sept 2026
- Humanity's Last Exam (AA): 46%9 Sept 2026
- IFBench: 69.2%9 Sept 2026
- AA-LCR: 81.7%9 Sept 2026
- MMMU-Pro: 81.9%9 Sept 2026
- and 67 more
- GPT-5.6 Terra
- AA Intelligence Index: 38.29 Sept 2026
- GPQA Diamond (AA): 90.8%9 Sept 2026
- Humanity's Last Exam (AA): 41.9%9 Sept 2026
- IFBench: 66.3%9 Sept 2026
- AA-LCR: 79%9 Sept 2026
- MMMU-Pro: 79.5%9 Sept 2026
- and 62 more
- GPT-5.5
- AA Intelligence Index: 38.69 Sept 2026
- GPQA Diamond (AA): 93.5%9 Sept 2026
- Humanity's Last Exam (AA): 45.8%9 Sept 2026
- IFBench: 75.9%9 Sept 2026
- AA-LCR: 84.3%9 Sept 2026
- MMMU-Pro: 79.9%9 Sept 2026
- and 58 more
- Claude Opus 4.5
- AA Intelligence Index: 29.19 Sept 2026
- GPQA Diamond (AA): 86.6%9 Sept 2026
- Humanity's Last Exam (AA): 30.1%9 Sept 2026
- IFBench: 58%9 Sept 2026
- AA-LCR: 77.3%9 Sept 2026
- MMMU-Pro: 74.1%9 Sept 2026
- and 48 more
- GPT-5
- AA Intelligence Index: 239 Sept 2026
- GPQA Diamond (AA): 85.4%9 Sept 2026
- Humanity's Last Exam (AA): 28.5%9 Sept 2026
- IFBench: 73.1%9 Sept 2026
- AA-LCR: 78.2%9 Sept 2026
- MMMU-Pro: 74.2%9 Sept 2026
- and 48 more
- GPT-5.6 Luna
- AA Intelligence Index: 34.89 Sept 2026
- GPQA Diamond (AA): 89.5%9 Sept 2026
- Humanity's Last Exam (AA): 37%9 Sept 2026
- AA-LCR: 81.7%9 Sept 2026
- MMMU-Pro: 78.6%9 Sept 2026
- SciCode (AA): 50.5%9 Sept 2026
- and 47 more
- GPT-6 Astra
- LMArena Agent: 12.59 Sept 2026
- Terminal-Bench 2.1 (Vals): 87.3%9 Sept 2026
- Vals Index: 66.69 Sept 2026
- Terminal-Bench: 58.2%9 Sept 2026
- AA Intelligence Index: 52.89 Sept 2026
- GPQA Diamond (AA): 96.1%9 Sept 2026
- and 46 more
- Claude Opus 5
- AA Intelligence Index: 50.79 Sept 2026
- GPQA Diamond (AA): 93.2%9 Sept 2026
- Humanity's Last Exam (AA): 54.9%9 Sept 2026
- AA-LCR: 79.3%9 Sept 2026
- MMMU-Pro: 84.7%9 Sept 2026
- SciCode (AA): 56.4%9 Sept 2026
- and 43 more
- GPT-5.4
- AA Intelligence Index: 399 Sept 2026
- GPQA Diamond (AA): 92%9 Sept 2026
- Humanity's Last Exam (AA): 43.7%9 Sept 2026
- IFBench: 74%9 Sept 2026
- AA-LCR: 82%9 Sept 2026
- MMMU-Pro: 78.4%9 Sept 2026
- and 41 more
- GPT-5.2
- AA Intelligence Index: 30.59 Sept 2026
- GPQA Diamond (AA): 90.3%9 Sept 2026
- Humanity's Last Exam (AA): 37.7%9 Sept 2026
- IFBench: 75.4%9 Sept 2026
- AA-LCR: 82.7%9 Sept 2026
- τ²-Bench Telecom (AA): 84.8%9 Sept 2026
- and 40 more
- Claude Sonnet 4
- AA Intelligence Index: 18.99 Sept 2026
- GPQA Diamond (AA): 77.7%9 Sept 2026
- Humanity's Last Exam (AA): 10.7%9 Sept 2026
- IFBench: 54.7%9 Sept 2026
- AA-LCR: 70.3%9 Sept 2026
- MMMU-Pro: 61.8%9 Sept 2026
- and 40 more
- Claude Sonnet 4.5
- AA Intelligence Index: 21.29 Sept 2026
- GPQA Diamond (AA): 83.4%9 Sept 2026
- Humanity's Last Exam (AA): 17.8%9 Sept 2026
- IFBench: 57.3%9 Sept 2026
- AA-LCR: 72.3%9 Sept 2026
- MMMU-Pro: 68.7%9 Sept 2026
- and 40 more
- Qwen3 Max
- AA Intelligence Index: 21.39 Sept 2026
- GPQA Diamond (AA): 86.1%9 Sept 2026
- Humanity's Last Exam (AA): 28%9 Sept 2026
- IFBench: 70.8%9 Sept 2026
- AA-LCR: 74.3%9 Sept 2026
- τ²-Bench Telecom (AA): 83.6%9 Sept 2026
- and 39 more
- Gemini 3.5 Flash
- AA Intelligence Index: 33.69 Sept 2026
- GPQA Diamond (AA): 92.1%9 Sept 2026
- Humanity's Last Exam (AA): 41.3%9 Sept 2026
- IFBench: 74.6%9 Sept 2026
- AA-LCR: 74.3%9 Sept 2026
- MMMU-Pro: 83.9%9 Sept 2026
- and 38 more
- Grok 4.20
- AA Intelligence Index: 25.79 Sept 2026
- GPQA Diamond (AA): 91.1%9 Sept 2026
- Humanity's Last Exam (AA): 34.5%9 Sept 2026
- IFBench: 82.9%9 Sept 2026
- AA-LCR: 69%9 Sept 2026
- MMMU-Pro: 74.6%9 Sept 2026
- and 38 more
- GPT-5 mini
- AA Intelligence Index: 17.49 Sept 2026
- GPQA Diamond (AA): 82.8%9 Sept 2026
- Humanity's Last Exam (AA): 21.5%9 Sept 2026
- IFBench: 75.4%9 Sept 2026
- AA-LCR: 72.3%9 Sept 2026
- MMMU-Pro: 70.1%9 Sept 2026
- and 37 more
- Claude Fable 5.1
- AA Intelligence Index: 53.29 Sept 2026
- GPQA Diamond (AA): 93.4%9 Sept 2026
- Humanity's Last Exam (AA): 58.7%9 Sept 2026
- AA-LCR: 83%9 Sept 2026
- SciCode (AA): 60.9%9 Sept 2026
- AA-Omniscience: 42.49 Sept 2026
- and 36 more
- Grok 4.1
- AA Intelligence Index: 20.49 Sept 2026
- GPQA Diamond (AA): 85.3%9 Sept 2026
- Humanity's Last Exam (AA): 19.3%9 Sept 2026
- IFBench: 52.7%9 Sept 2026
- AA-LCR: 74%9 Sept 2026
- MMMU-Pro: 63.3%9 Sept 2026
- and 36 more
- DeepSeek V4 Pro
- AA Intelligence Index: 30.19 Sept 2026
- GPQA Diamond (AA): 90.5%9 Sept 2026
- Humanity's Last Exam (AA): 35.2%9 Sept 2026
- IFBench: 71.3%9 Sept 2026
- AA-LCR: 70.3%9 Sept 2026
- τ²-Bench Telecom (AA): 94.2%9 Sept 2026
- and 35 more
- Grok 4 Fast
- AA Intelligence Index: 17.99 Sept 2026
- GPQA Diamond (AA): 84.8%9 Sept 2026
- Humanity's Last Exam (AA): 19.1%9 Sept 2026
- IFBench: 50.5%9 Sept 2026
- AA-LCR: 73.7%9 Sept 2026
- MMMU-Pro: 61.8%9 Sept 2026
- and 35 more
- GPT-5.4 mini
- AA Intelligence Index: 24.69 Sept 2026
- GPQA Diamond (AA): 87.5%9 Sept 2026
- Humanity's Last Exam (AA): 28.1%9 Sept 2026
- IFBench: 73.3%9 Sept 2026
- AA-LCR: 77%9 Sept 2026
- MMMU-Pro: 73.3%9 Sept 2026
- and 34 more
- GPT-5.4 nano
- AA Intelligence Index: 21.29 Sept 2026
- GPQA Diamond (AA): 81.7%9 Sept 2026
- Humanity's Last Exam (AA): 28.3%9 Sept 2026
- IFBench: 75.9%9 Sept 2026
- AA-LCR: 76.7%9 Sept 2026
- MMMU-Pro: 65.4%9 Sept 2026
- and 34 more
- DeepSeek V4 Flash
- AA Intelligence Index: 24.89 Sept 2026
- GPQA Diamond (AA): 86.7%9 Sept 2026
- Humanity's Last Exam (AA): 30.3%9 Sept 2026
- IFBench: 73.5%9 Sept 2026
- AA-LCR: 72%9 Sept 2026
- SciCode (AA): 40.2%9 Sept 2026
- and 33 more
- Gemini 3.8 Flash
- AA Intelligence Index: 409 Sept 2026
- GPQA Diamond (AA): 93.5%9 Sept 2026
- Humanity's Last Exam (AA): 42.1%9 Sept 2026
- AA-LCR: 84%9 Sept 2026
- MMMU-Pro: 84.2%9 Sept 2026
- SciCode (AA): 55.1%9 Sept 2026
- and 32 more
Who is due
Expected windows from each provider’s release cadence over the last two years. An estimate from dates only, not from announcements.
| Provider | Last release | Cadence | Expected window |
|---|---|---|---|
| Amazon | 2 Dec 2025 | every 34 days | 22 Dec 2025 to 18 Jan 2026overdue |
| Xiaomi | 23 Apr 2026 | every 35 days | 14 May 2026 to 11 Jun 2026overdue |
 Mistral AI | 13 Jun 2026 | every 46 days | 10 Jul 2026 to 16 Aug 2026overdue |
| Cohere | 9 Jun 2026 | every 87 days | 31 Jul 2026 to 8 Oct 2026 |
| MiniMax | 1 Jun 2026 | every 108 days | 4 Aug 2026 to 30 Oct 2026 |
| Unknown | 3 Aug 2026 | every 36 days | 24 Aug 2026 to 22 Sept 2026 |
| NVIDIA | 11 Aug 2026 | every 26 days | 26 Aug 2026 to 16 Sept 2026 |
| Alibaba | 26 Aug 2026 | every 6 days | 29 Aug 2026 to 3 Sept 2026overdue |
 Moonshot AI | 16 Jul 2026 | every 82 days | 3 Sept 2026 to 7 Nov 2026 |
| Zhipu AI | 20 Aug 2026 | every 28 days | 5 Sept 2026 to 28 Sept 2026 |
| xAI | 12 Aug 2026 | every 42 days | 6 Sept 2026 to 9 Oct 2026 |
| Google | 2 Sept 2026 | every 20 days | 14 Sept 2026 to 30 Sept 2026 |
| OpenAI | 3 Sept 2026 | every 26 days | 18 Sept 2026 to 9 Oct 2026 |
| Meta | 2 Sept 2026 | every 27 days | 18 Sept 2026 to 9 Oct 2026 |
| Anthropic | 1 Sept 2026 | every 40 days | 25 Sept 2026 to 27 Oct 2026 |
 DeepSeek | 21 Aug 2026 | every 84 days | 10 Oct 2026 to 16 Dec 2026 |
| IBM | 25 Aug 2026 | every 118 days | 3 Nov 2026 to 6 Feb 2027 |
| Upstage | 6 Aug 2026 | every 191 days | 28 Nov 2026 to 30 Apr 2027 |
| StepFun | 29 May 2026 | every 406 days | 27 Jan 2027 to 18 Dec 2027 |