[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"models":3},[4,58,122,171,211,272,305,350,402,429,457,495,537,584,624,684,724,748,792,837,873,902,959,999,1027,1055,1076,1106,1144,1194,1221,1259,1313],{"id":5,"name":6,"slug":7,"provider":8,"family":9,"variant":10,"description":11,"url":12,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":14,"outputLimit":14,"modalities":15,"modalitiesOut":19,"inputPricePerMTok":20,"outputPricePerMTok":21,"cacheReadPricePerMTok":22,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":23,"lastUpdated":26,"knowledge":27,"status":29,"reasoning":30,"reasoningControl":31,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":33,"openrouterUrl":34,"hf":-1,"company":-1,"articles":35,"createdAt":57},"entity_01kzvnwvgdesp9z6m0yezqs963","Grok 4.6","grok-4-6","xAI","Grok","4.6","xAI's frontier Grok model, tuned for long-running agents, coding, knowledge work, and visual projects.","https:\u002F\u002Fx.ai\u002Fnews\u002Fgrok-4-6",false,500000,[16,17,18],"text","image","pdf",[16],2,6,0.5,{"value":24,"precision":25},"2026-08-12","day",{"value":24,"precision":25},{"value":28,"precision":25},"2026-02-01","ga",true,[32],"effort","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fxai\u002Fgrok-4.6\u002F","https:\u002F\u002Fopenrouter.ai\u002Fx-ai\u002Fgrok-4.6",[36,49],{"id":37,"name":38,"slug":39,"summary":40,"url":41,"kind":42,"platform":43,"author":44,"authorHandle":45,"publisher":46,"publishedAt":24,"about":47,"writtenBy":48,"createdAt":-1},"entity_01kzvpwb44esgb3hrkgrdryghw","Grok 4.6 – A field guide","grok-4-6-a-field-guide","Grok 4.6 is out! I've used it for a few weeks as my daily driver across the normal mix of coding and knowledge work, and built a few projects with it specifically to push on where it holds up.","https:\u002F\u002Fx.com\u002Fericzakariasson\u002Fstatus\u002F2087566447178547494","article","x","eric zakariasson","ericzakariasson","X",[],[],{"id":50,"name":51,"slug":52,"summary":53,"url":12,"kind":42,"platform":54,"author":8,"authorHandle":-1,"publisher":8,"publishedAt":24,"about":55,"writtenBy":56,"createdAt":-1},"entity_01kzvngmfgesp9z6kwae8h36zq","Introducing Grok 4.6","introducing-grok-4-6","Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work.","web",[],[],1786561588,{"id":59,"name":60,"slug":61,"provider":62,"family":63,"variant":64,"description":65,"url":66,"openWeights":30,"license":67,"parameters":68,"contextWindow":69,"outputLimit":-1,"modalities":70,"modalitiesOut":71,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":72,"releasedAt":73,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":13,"reasoningControl":74,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":75,"company":-1,"articles":111,"createdAt":121},"entity_01kzw21fbye068x2smxr8rna56","LFM2.5-VL-3B","lfm2-5-vl-3b","Liquid AI","LFM2.5","VL-3B","LFM2.5-VL-3B is the multimodal variant of LFM2.5, pairing the LFM2.5-2.6B backbone with a SigLIP2 NaFlex 400M vision encoder. It answers directly rather than reasoning, and adds screen understanding, grounding and function calling for on-device use.","https:\u002F\u002Fhuggingface.co\u002FLiquidAI\u002FLFM2.5-VL-3B","LFM Open License v1.0","3.1B",32768,[16,17],[16],{"value":24,"precision":25},{"value":24,"precision":25},[],{"fullName":76,"url":66,"downloads":77,"likes":78,"license":79,"pipelineTag":80,"tags":81,"lastModified":110},"LiquidAI\u002FLFM2.5-VL-3B",0,80,"other","image-text-to-text",[82,83,84,80,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109],"transformers","safetensors","lfm2_vl","liquid","lfm2.5","edge","conversational","custom_code","ar","zh","en","fr","de","hi","id","it","ja","ko","pl","pt","ru","es","th","vi","arxiv:2305.03393","license:other","endpoints_compatible","region:us","2026-08-12T15:36:46.000Z",[112],{"id":113,"name":114,"slug":115,"summary":116,"url":117,"kind":118,"platform":54,"author":62,"authorHandle":-1,"publisher":62,"publishedAt":24,"about":119,"writtenBy":120,"createdAt":-1},"entity_01kzw21yh9exwtra57715r4hz2","LFM2.5-VL-3B: A Better and Faster Vision-Language Model for the Edge","lfm2-5-vl-3b-a-better-and-faster-vision-language-model-for-the-edge","Liquid AI's most capable vision-language model: 3.1B params, direct answers instead of reasoning, with big gains in screen understanding, grounding and function calling. Open weights on Hugging Face.","https:\u002F\u002Fwww.liquid.ai\u002Fblog\u002Flfm2-5-vl-3b","announcement",[],[],1786574323,{"id":123,"name":124,"slug":125,"provider":126,"family":126,"variant":127,"description":128,"url":129,"openWeights":30,"license":130,"parameters":-1,"contextWindow":-1,"outputLimit":-1,"modalities":131,"modalitiesOut":134,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":135,"releasedAt":137,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":138,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":139,"company":163,"articles":169,"createdAt":170},"entity_01kzs9vthtecwbv562bdd7v182","LTX-2.5","ltx-2-5","LTX","2.5","LTX-2.5 generates multi-shot scenes in one pass, edits real footage, and exports cinema-grade EXR. Open weights you can fine-tune and run on your hardware.","https:\u002F\u002Fltx.io\u002Fmodel\u002Fltx-2-5","LTX-2 Community License Agreement",[16,17,132,133],"video","audio",[132,133],{"value":136,"precision":25},"2026-08-11",{"value":136,"precision":25},[],{"fullName":140,"url":141,"downloads":142,"likes":143,"license":79,"pipelineTag":144,"tags":145,"lastModified":162},"Lightricks\u002FLTX-2.5","https:\u002F\u002Fhuggingface.co\u002FLightricks\u002FLTX-2.5",39,551,"image-to-video",[146,144,147,148,149,150,151,152,153,154,155,156,157,158,159,160,92,94,103,93,98,99,91,97,101,161,107,109],"diffusion-single-file","text-to-video","video-to-video","image-text-to-video","audio-to-video","text-to-audio","video-to-audio","audio-to-audio","text-to-audio-video","image-to-audio-video","image-text-to-audio-video","ltx-video","lightricks","comfyui","ltx-2.5","arxiv:2601.03233","2026-08-12T14:20:43.000Z",{"id":164,"name":126,"slug":165,"url":166,"bio":167,"githubOrg":168},"entity_01kzs9wvrqextv2kqpwqa099ym","ltx","https:\u002F\u002Fltx.io","LTX builds open-weight diffusion transformer foundation models for multimodal video, audio, and world simulation.","Lightricks",[],1786481863,{"id":172,"name":173,"slug":174,"provider":175,"family":176,"variant":177,"description":178,"url":179,"openWeights":30,"license":180,"parameters":181,"contextWindow":182,"outputLimit":-1,"modalities":183,"modalitiesOut":184,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":185,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":186,"toolCall":30,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":187,"company":204,"articles":209,"createdAt":210},"entity_01kzs5yg3nfantmwktneyeq6y9","Needle 2","needle-2","Cactus Compute","Needle","2","An open 45M-parameter model for tool calling, device use, and structured extraction. Needle 2 runs as a 14 MB binary in 28 MB of session RAM.","https:\u002F\u002Fcactuscompute.com\u002Fneedle","Apache-2.0","45M",256,[16],[16],{"value":136,"precision":25},[],{"fullName":188,"url":189,"downloads":77,"likes":190,"license":191,"pipelineTag":192,"tags":193,"lastModified":203},"Cactus-Compute\u002Fneedle2","https:\u002F\u002Fhuggingface.co\u002FCactus-Compute\u002Fneedle2",71,"apache-2.0","text-generation",[194,195,196,197,198,87,199,200,192,201,202,109],"cactus-needle","needle","tool-calling","function-calling","on-device","quantization","webassembly","arxiv:2607.18363","license:apache-2.0","2026-08-12T19:45:19.000Z",{"id":205,"name":175,"slug":206,"url":207,"bio":208,"githubOrg":206},"entity_01kzs5zee0fantmwn5ff5a5en7","cactus-compute","https:\u002F\u002Fcactuscompute.com","On-device AI with cloud fallback. Cactus post-trains models to know when they are wrong and hand off to frontier cloud models — cutting inference costs up to 5x.",[],1786477756,{"id":212,"name":213,"slug":214,"provider":215,"family":216,"variant":217,"description":218,"url":219,"openWeights":30,"license":220,"parameters":221,"contextWindow":222,"outputLimit":222,"modalities":223,"modalitiesOut":224,"inputPricePerMTok":77,"outputPricePerMTok":77,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":225,"lastUpdated":226,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":227,"toolCall":30,"attachment":13,"aiSdkId":229,"modelsDevUrl":230,"openrouterUrl":231,"hf":232,"company":246,"articles":250,"createdAt":271},"entity_01kzrx0dn0e8h9epq84mme2d6d","Nemotron 3.5 Lightning","nemotron-3-5-lightning","NVIDIA","Nemotron","3.5 Lightning","A customizable open 30B MoE model with 3B active parameters, providing optimal high-volume execution for autonomous agents.","https:\u002F\u002Fhuggingface.co\u002Fnvidia\u002FNVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","OpenMDW-1.1","30B total, 3B active (MoE)",262144,[16],[16],{"value":136,"precision":25},{"value":136,"precision":25},[228],"toggle","nemotron-3.5-lightning-30b-a3b","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fnvidia\u002Fnemotron-3.5-lightning-30b-a3b\u002F","https:\u002F\u002Fopenrouter.ai\u002Fnvidia\u002Fnemotron-3.5-lightning",{"fullName":233,"url":219,"downloads":234,"likes":235,"license":79,"pipelineTag":192,"tags":236,"lastModified":245},"nvidia\u002FNVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16",15740,109,[82,83,237,192,238,239,240,88,92,103,93,94,97,98,241,242,107,243,108,244,109],"nemotron_h","nvidia","pytorch","nemotron-3.5","dataset:nvidia\u002Fnemotron-post-training-v3","dataset:nvidia\u002Fnemotron-pre-training-datasets","eval-results","deploy:azure","2026-08-12T17:21:44.000Z",{"id":247,"name":215,"slug":238,"url":248,"bio":249,"githubOrg":215},"entity_01kzrxjf0ze8h9epsbfe0560wm","https:\u002F\u002Fwww.nvidia.com","NVIDIA invents the GPU and drives advances in AI, HPC, gaming, creative design, autonomous vehicles, and robotics.",[251,261],{"id":252,"name":253,"slug":254,"summary":255,"url":256,"kind":42,"platform":43,"author":257,"authorHandle":258,"publisher":46,"publishedAt":24,"about":259,"writtenBy":260,"createdAt":-1},"entity_01kzwdw503esnadb14a5zqk83b","Nemotron 3.5 Lightning 30B-A3B: Behavioral Analysis","nemotron-3-5-lightning-30b-a3b-behavioral-analysis","Pre-note. This is a behavioral audit, not a vendor takedown.","https:\u002F\u002Fx.com\u002Fno_stp_on_snek\u002Fstatus\u002F2087625566484562424","Tom Turney","no_stp_on_snek",[],[],{"id":262,"name":263,"slug":264,"summary":265,"url":266,"kind":118,"platform":54,"author":267,"authorHandle":-1,"publisher":268,"publishedAt":136,"about":269,"writtenBy":270,"createdAt":-1},"entity_01kzrx0zhne8h9eprh4wwvwh4b","NVIDIA Nemotron 3.5 Lightning Delivers Fast, Accurate Specialized Task Execution for Long-Running Agents","nvidia-nemotron-3-5-lightning-delivers-fast-accurate-specialized-task-execution","NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for the execution layer of always-on agents, designed for harnesses like OpenClaw and Hermes Agent.","https:\u002F\u002Fdeveloper.nvidia.com\u002Fblog\u002Fnvidia-nemotron-3-5-lightning-delivers-fast-accurate-specialized-task-execution-for-long-running-agents\u002F","Chris Alexiuk and Chintan Patel","NVIDIA Technical Blog",[],[],1786468382,{"id":273,"name":274,"slug":275,"provider":276,"family":277,"variant":278,"description":279,"url":280,"openWeights":30,"license":281,"parameters":282,"contextWindow":222,"outputLimit":69,"modalities":283,"modalitiesOut":284,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":285,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":287,"toolCall":30,"attachment":-1,"aiSdkId":288,"modelsDevUrl":-1,"openrouterUrl":289,"hf":290,"company":298,"articles":303,"createdAt":304},"entity_01kzs9848nextv2knyrtd8120z","Ling 3.0 Tiny","ling-3-0-tiny","inclusionAI","Ling","3.0 Tiny","We are introducing Ling-3.0-tiny, a lightweight hybrid reasoning MoE model with 7.9B total parameters and only 1.3B activated parameters per token. It is designed to deliver strong reasoning and agentic capabilities at low inference cost, making advanced model capabilities more accessible for local and resource-constrained deployment.","https:\u002F\u002Fhuggingface.co\u002FinclusionAI\u002FLing-3.0-tiny","MIT","7.9B total, 1.3B active (MoE)",[16],[16],{"value":286,"precision":25},"2026-08-10",[228],"ling-3.0-tiny","https:\u002F\u002Fopenrouter.ai\u002Finclusionai\u002Fling-3.0-tiny",{"fullName":291,"url":280,"downloads":77,"likes":292,"license":293,"pipelineTag":-1,"tags":294,"lastModified":297},"inclusionAI\u002FLing-3.0-tiny",184,"mit",[83,295,89,296,109],"bailing_hybrid","license:mit","2026-08-11T07:17:36.000Z",{"id":299,"name":276,"slug":300,"url":301,"bio":302,"githubOrg":276},"entity_01kzs9fvntextv2kpmyhh4cv9c","inclusion-ai","https:\u002F\u002Fwww.inclusion-ai.org","inclusionAI (IAI) envisions AGI as humanity's shared milestone, not a privileged asset",[],1786481217,{"id":306,"name":307,"slug":308,"provider":309,"family":310,"variant":311,"description":312,"url":313,"openWeights":30,"license":180,"parameters":314,"contextWindow":315,"outputLimit":-1,"modalities":316,"modalitiesOut":317,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":318,"releasedAt":319,"lastUpdated":-1,"knowledge":320,"status":29,"reasoning":30,"reasoningControl":322,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":323,"hf":324,"company":332,"articles":338,"createdAt":349},"entity_01kzpbt0v2fvn8q02mecr6ffbz","Muse Glimmer 30B","muse-glimmer-30b","Meta","Muse","Glimmer 30B","Muse Glimmer is a 30-billion-parameter causal language model with a dedicated perception encoder, distilled from Muse Spark and purpose-built for autonomous agentic tasks on consumer hardware. The model integrates multi-step reasoning, reliable tool use, multimodal understanding, and failure recovery into a single model that runs locally without requiring cloud infrastructure or network access.","https:\u002F\u002Fhuggingface.co\u002Fmeta-models\u002FMuse-Glimmer-30B","29.6B",131072,[16,17],[16],{"value":286,"precision":25},{"value":286,"precision":25},{"value":321,"precision":25},"2026-01-04",[32],"https:\u002F\u002Fopenrouter.ai\u002Fmeta\u002Fmuse-glimmer-30b",{"fullName":325,"url":313,"downloads":77,"likes":326,"license":191,"pipelineTag":80,"tags":327,"lastModified":331},"meta-models\u002FMuse-Glimmer-30B",1272,[82,83,328,80,88,329,330,202,243,108,109],"muse_glimmer","arxiv:2504.13181","arxiv:2602.06036","2026-08-11T19:23:35.000Z",{"id":333,"name":309,"slug":334,"url":335,"bio":336,"githubOrg":337},"entity_01kz9zhkxxfztsw2p4fej7zagt","meta","https:\u002F\u002Fai.meta.com","Explore AI at Meta. Use Meta AI to answer questions, create images, and complete tasks for free. Plus AI research, models, and tools for developers.","facebook",[339],{"id":340,"name":341,"slug":342,"summary":343,"url":344,"kind":118,"platform":54,"author":345,"authorHandle":-1,"publisher":346,"publishedAt":286,"about":347,"writtenBy":348,"createdAt":-1},"entity_01kzpbxhygfvn8q042sxw8nm0n","Introducing Muse Glimmer: An Open Agentic Model That Runs on Your Device","introducing-muse-glimmer-an-open-agentic-model-that-runs-on-your-device","Muse Glimmer is a 30-billion-parameter open agentic model from Meta Superintelligence Labs, optimized for always-on local workflows on consumer hardware.","https:\u002F\u002Fresearch.meta.ai\u002Fblog\u002Fintroducing-muse-glimmer-open-agentic-model","Meta Superintelligence Labs","Meta AI Research",[],[],1786383237,{"id":351,"name":352,"slug":353,"provider":354,"family":355,"variant":356,"description":357,"url":358,"openWeights":30,"license":359,"parameters":356,"contextWindow":360,"outputLimit":-1,"modalities":361,"modalitiesOut":362,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":363,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":365,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":366,"company":386,"articles":391,"createdAt":401},"entity_01kzpfy8bffvn8q0j1xqte9b8q","TwiL-LM 1.7B","twil-lm-1-7b","webAI","TwiL-LM","1.7B","A LoRA adapter over SmolLM2-1.7B-Instruct for formal logic — FOL translation, entailment, semantic parsing and Lean assistance — quantized to a 1.06 GB download that runs on a phone.","https:\u002F\u002Fhuggingface.co\u002FwebAI-Official\u002FTwIL-LM","webAI Non-Commercial License v1.0",8192,[16],[16],{"value":364,"precision":25},"2026-08-07",[],{"fullName":367,"url":358,"downloads":368,"likes":369,"license":79,"pipelineTag":192,"tags":370,"lastModified":385},"webAI-Official\u002FTwIL-LM",519,11,[82,83,371,372,192,373,374,375,376,377,378,379,380,381,88,92,382,383,107,384,108,109],"gguf","llama","formal-logic","reasoning","lora","model-merging","wise-ft","reinforcement-learning","grpo","smollm2","twil-lm","base_model:HuggingFaceTB\u002FSmolLM2-1.7B-Instruct","base_model:adapter:HuggingFaceTB\u002FSmolLM2-1.7B-Instruct","text-generation-inference","2026-08-12T21:22:59.000Z",{"id":387,"name":354,"slug":388,"url":389,"bio":390,"githubOrg":-1},"entity_01kz4rc7egfvmtc6gnd5zdmkwr","webai","https:\u002F\u002Fwww.webai.com\u002F","Build specialized AI that runs where your data lives. webAI is a sovereign AI platform for creating, owning, and connecting specialized intelligence across networks.",[392],{"id":393,"name":394,"slug":395,"summary":396,"url":397,"kind":42,"platform":54,"author":398,"authorHandle":-1,"publisher":354,"publishedAt":364,"about":399,"writtenBy":400,"createdAt":-1},"entity_01kzpfpt6xfvn8q0hkfg68smnm","webAI Releases TwiL-LM, a Family of Formal-Logic Models That Outreason a 120B Model and Run on an iPhone","webai-releases-twil-lm-a-family-of-formal-logic-models-that-outreason-a-120b","Built for compliance rules, contract logic, and research reasoning, the 3B model beats OpenAI's open-weights gpt-oss-120b, a model 40× its size, on four of five formal-reasoning benchmarks, while the 1-gigabyte 1.7B variant outperforms every sub-2B model webAI evaluated.","https:\u002F\u002Fwww.webai.com\u002Fblog\u002Fwebai-releases-twil-lm-a-family-of-formal-logic-models-that-outreason-a-120b-model-and-run-on-an-iphone","webAI Team",[],[],1786387571,{"id":403,"name":404,"slug":405,"provider":354,"family":355,"variant":406,"description":407,"url":408,"openWeights":30,"license":359,"parameters":406,"contextWindow":409,"outputLimit":-1,"modalities":410,"modalitiesOut":411,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":412,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":413,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":414,"company":423,"articles":424,"createdAt":428},"entity_01kzpfya86fvn8q0kajqww0x79","TwiL-LM3","twil-lm3","3B","A 3B formal-logic reasoning model built from SmolLM3-3B by LoRA fine-tuning, checkpoint fusion, WiSE-FT interpolation and entropy-weighted GRPO, which beats gpt-oss-120b on four of five lanes of webAI's formal-reasoning suite.","https:\u002F\u002Fhuggingface.co\u002FwebAI-Official\u002FTwIL-LM3",65536,[16],[16],{"value":364,"precision":25},[],{"fullName":415,"url":408,"downloads":416,"likes":417,"license":79,"pipelineTag":192,"tags":418,"lastModified":422},"webAI-Official\u002FTwIL-LM3",289,45,[82,83,371,419,192,373,374,375,376,377,378,379,381,88,92,420,421,107,108,109],"smollm3","base_model:HuggingFaceTB\u002FSmolLM3-3B","base_model:adapter:HuggingFaceTB\u002FSmolLM3-3B","2026-08-12T21:22:50.000Z",{"id":387,"name":354,"slug":388,"url":389,"bio":390,"githubOrg":-1},[425],{"id":393,"name":394,"slug":395,"summary":396,"url":397,"kind":42,"platform":54,"author":398,"authorHandle":-1,"publisher":354,"publishedAt":364,"about":426,"writtenBy":427,"createdAt":-1},[],[],1786387572,{"id":430,"name":431,"slug":432,"provider":433,"family":434,"variant":435,"description":436,"url":437,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":-1,"outputLimit":-1,"modalities":438,"modalitiesOut":439,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":440,"releasedAt":442,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":444,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":-1,"company":445,"articles":450,"createdAt":456},"entity_01ky96h4hye4a93wg1ck39y56f","FLUX 3","flux-3","Black Forest Labs","FLUX","3","FLUX 3 is our new multimodal foundation model. It jointly learns from images, videos, and audio within a unified architecture, because what it needs to learn is not any one of these elements in isolation. Instead, a model must learn a representation of the world: how objects hold together, how things move, and how events sound.","https:\u002F\u002Fbfl.ai\u002Fblog\u002Fflux-3",[16,17,132,133],[132,133],{"value":441,"precision":25},"2026-07-23",{"value":443,"precision":25},"2026-08-04",[],{"id":446,"name":433,"slug":447,"url":448,"bio":449,"githubOrg":447},"entity_01kyn5fy1ze4frj5d18cwjkww6","black-forest-labs","https:\u002F\u002Fbfl.ai","Black Forest Labs is building visual intelligence: models that understand, reason, and act in the world. Use FLUX models via our API.",[451],{"id":452,"name":453,"slug":432,"summary":436,"url":437,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":433,"publishedAt":441,"about":454,"writtenBy":455,"createdAt":-1},"entity_01ky96hg83e4a93wgdm2d8m9nq","FLUX 3 - Real World Models: Towards Multimodal Flow Models as the Backbone of Visual Intelligence",[],[],1784867754,{"id":458,"name":459,"slug":460,"provider":62,"family":63,"variant":461,"description":462,"url":463,"openWeights":30,"license":67,"parameters":461,"contextWindow":315,"outputLimit":-1,"modalities":464,"modalitiesOut":465,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":466,"releasedAt":467,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":13,"reasoningControl":468,"toolCall":30,"attachment":13,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":469,"company":479,"articles":485,"createdAt":494},"entity_01kz6swcenf6d8hgrwc43tq280","LFM2.5-2.6B","lfm2-5-2-6b","2.6B","LFM2.5-2.6B is part of LFM2.5, a family of hybrid models designed for on-device deployment. It builds on the LFM2 architecture with a 128K context window and agentic post-training.","https:\u002F\u002Fhuggingface.co\u002FLiquidAI\u002FLFM2.5-2.6B",[16],[16],{"value":443,"precision":25},{"value":443,"precision":25},[],{"fullName":470,"url":463,"downloads":471,"likes":472,"license":79,"pipelineTag":192,"tags":473,"lastModified":478},"LiquidAI\u002FLFM2.5-2.6B",93668,579,[82,83,474,192,85,86,87,88,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,475,476,477,107,243,108,109],"lfm2","arxiv:2511.23404","base_model:LiquidAI\u002FLFM2.5-2.6B-Base","base_model:finetune:LiquidAI\u002FLFM2.5-2.6B-Base","2026-08-07T10:30:38.000Z",{"id":480,"name":62,"slug":481,"url":482,"bio":483,"githubOrg":484},"entity_01kz6syj0vfk0v7j0pn2gds1v2","liquid-ai","https:\u002F\u002Fwww.liquid.ai","Liquid AI is an efficiency-first foundation model company. We build highly capable, compute-optimized models that bring intelligence to any device and medium of choice.","Liquid4All",[486],{"id":487,"name":488,"slug":489,"summary":490,"url":491,"kind":118,"platform":54,"author":62,"authorHandle":-1,"publisher":62,"publishedAt":443,"about":492,"writtenBy":493,"createdAt":-1},"entity_01kz6we832e05brtr0t5te8ran","LFM2.5-2.6B: Deploy Agents Everywhere","lfm2-5-2-6b-deploy-agents-everywhere","LFM2.5-2.6B is an on-device agentic model that plans, calls tools, and runs multi-step tasks at 220 tok\u002Fs in under 2.5 GB. Open weights on Hugging Face.","https:\u002F\u002Fwww.liquid.ai\u002Fblog\u002Flfm2-5-2-6b",[],[],1785861124,{"id":496,"name":497,"slug":498,"provider":499,"family":497,"variant":500,"description":501,"url":502,"openWeights":30,"license":180,"parameters":406,"contextWindow":222,"outputLimit":-1,"modalities":503,"modalitiesOut":504,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":505,"releasedAt":506,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":13,"reasoningControl":507,"toolCall":13,"attachment":30,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":508,"company":521,"articles":527,"createdAt":536},"entity_01kz6y577secrbyec0aa760r3x","Shieldstral","shieldstral","Mistral AI","1.0 3B","A 3B open-weights, policy-adaptive multimodal safety classifier that matches models up to 7x its size on text safety and sets a new state of the art on multimodal moderation.","https:\u002F\u002Fhuggingface.co\u002Fmistralai\u002FShieldstral-1.0-3B",[16,17],[16],{"value":443,"precision":25},{"value":443,"precision":25},[],{"fullName":509,"url":502,"downloads":510,"likes":511,"license":191,"pipelineTag":-1,"tags":512,"lastModified":520},"mistralai\u002FShieldstral-1.0-3B",6769,235,[513,83,514,515,92,93,103,94,97,101,516,91,98,99,90,102,517,518,519,202,109],"vllm","mistral3","mistral-common","nl","arxiv:2607.25857","base_model:mistralai\u002FMinistral-3-3B-Base-2512","base_model:finetune:mistralai\u002FMinistral-3-3B-Base-2512","2026-08-05T10:06:52.000Z",{"id":522,"name":499,"slug":523,"url":524,"bio":525,"githubOrg":526},"entity_01kz6y3gbpeh3r944y2ehaw1jh","mistral-ai","https:\u002F\u002Fmistral.ai","The most powerful AI platform for enterprises. Customize, fine-tune, and deploy AI assistants, autonomous agents, and multimodal AI with open models.","mistralai",[528],{"id":529,"name":530,"slug":531,"summary":532,"url":533,"kind":118,"platform":54,"author":499,"authorHandle":-1,"publisher":499,"publishedAt":443,"about":534,"writtenBy":535,"createdAt":-1},"entity_01kz6xz65xecrbyebjbymavnas","Introducing Shieldstral.","introducing-shieldstral","Shieldstral introduces a 3B open-weights multimodal safety classifier that outperforms models up to 7x its size.","https:\u002F\u002Fmistral.ai\u002Fnews\u002Fshieldstral\u002F",[],[],1785865608,{"id":538,"name":539,"slug":540,"provider":541,"family":541,"variant":542,"description":543,"url":544,"openWeights":30,"license":545,"parameters":546,"contextWindow":-1,"outputLimit":-1,"modalities":547,"modalitiesOut":548,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":549,"releasedAt":551,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":553,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":554,"company":568,"articles":574,"createdAt":583},"entity_01kz2yathse4br33n2kn6nkkfz","MiniMax H3","minimax-h3","MiniMax","H3","MiniMax H3 is a general-purpose, omni-modal generative system. It supports unified understanding of multimodal contexts composed of text, images, video, and audio, and can generate video with native stereo audio at resolutions up to 2K and durations of up to 15 seconds.","https:\u002F\u002Fhuggingface.co\u002FMiniMaxAI\u002FMiniMax-H3","MiniMax H3 Community License Agreement","33B",[16,17,132,133],[132,133],{"value":550,"precision":25},"2026-07-31",{"value":552,"precision":25},"2026-08-03",[],{"fullName":555,"url":544,"downloads":556,"likes":557,"license":79,"pipelineTag":149,"tags":558,"lastModified":567},"MiniMaxAI\u002FMiniMax-H3",83484,3702,[559,83,147,144,149,148,154,155,156,560,561,562,563,564,565,107,566,109],"diffusers","video-to-audio-video","audio-to-audio-video","audio-video-generation","multimodal","synchronized-audio-video","reference-to-audio-video","diffusers:MiniMaxH3ModularPipeline","2026-08-11T09:06:41.000Z",{"id":569,"name":541,"slug":570,"url":571,"bio":572,"githubOrg":573},"entity_01kz2y973hene8d7k5wn8bczwz","minimax","https:\u002F\u002Fwww.minimax.io","Building AGI with our mission Intelligence with Everyone. Global leader in multi-modal models and AI-native products with over 200 million users.","MiniMax-AI",[575],{"id":576,"name":577,"slug":578,"summary":579,"url":580,"kind":118,"platform":54,"author":541,"authorHandle":-1,"publisher":541,"publishedAt":550,"about":581,"writtenBy":582,"createdAt":-1},"entity_01kz2yhcgne4br33pmghxn59ha","MiniMax H3: An Open Model Breaking the Boundaries Between Tasks and Modalities","minimax-h3-an-open-model-breaking-the-boundaries-between-tasks-and-modalities","Today, we're officially launching MiniMax H3, a general-purpose omni-modal generation model. H3 can jointly understand multimodal contexts spanning text, images, video, and audio. It generates video with native stereo audio at up to 2K resolution and 15 seconds in length.","https:\u002F\u002Fwww.minimax.io\u002Fblog\u002Fminimax-h3",[],[],1785731574,{"id":585,"name":586,"slug":587,"provider":588,"family":589,"variant":590,"description":591,"url":592,"openWeights":30,"license":-1,"parameters":-1,"contextWindow":-1,"outputLimit":-1,"modalities":593,"modalitiesOut":594,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":595,"releasedAt":597,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":598,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":599,"openrouterUrl":600,"hf":601,"company":611,"articles":615,"createdAt":623},"entity_01kytdp3j0f22s8x568dssqppx","Inkling-Small","inkling-small","Thinking Machines Lab","Inkling","Small","A quarter the size of Inkling at comparable performance: a multimodal Mixture-of-Experts reasoner (276B total, 12B active) with controllable reasoning effort, fine-tunable on Tinker.","https:\u002F\u002Fthinkingmachines.ai\u002Fnews\u002Finkling-small\u002F",[],[],{"value":596,"precision":25},"2026-07-30",{"value":596,"precision":25},[],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fthinkingmachines\u002FInkling-Small\u002F","https:\u002F\u002Fopenrouter.ai\u002Fthinkingmachines\u002Finkling-small",{"fullName":602,"url":603,"downloads":604,"likes":605,"license":191,"pipelineTag":80,"tags":606,"lastModified":610},"thinkingmachines\u002FInkling-Small","https:\u002F\u002Fhuggingface.co\u002Fthinkingmachines\u002FInkling-Small",35390,354,[82,83,607,80,88,608,609,202,243,108,109],"inkling_mm_model","audio-text-to-text","moe","2026-07-31T00:39:44.000Z",{"id":612,"name":588,"slug":613,"url":614,"bio":-1,"githubOrg":613},"entity_01kyn5g3kce4frj5e6rt2c3rpm","thinking-machines-lab","https:\u002F\u002Fthinkingmachines.ai",[616],{"id":617,"name":618,"slug":619,"summary":620,"url":592,"kind":42,"platform":54,"author":588,"authorHandle":-1,"publisher":588,"publishedAt":596,"about":621,"writtenBy":622,"createdAt":-1},"entity_01kytdkp68e8jsra7tjvswdf63","Introducing Inkling-Small","introducing-inkling-small","An open-weights model that matches Inkling at a quarter of the size: multimodal, Mixture-of-Experts, with controllable reasoning effort. Fine-tune it on Tinker.",[],[],1785445682,{"id":625,"name":626,"slug":627,"provider":628,"family":629,"variant":630,"description":631,"url":632,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":633,"outputLimit":634,"modalities":635,"modalitiesOut":636,"inputPricePerMTok":637,"outputPricePerMTok":638,"cacheReadPricePerMTok":22,"cacheWritePricePerMTok":639,"announcedAt":-1,"releasedAt":640,"lastUpdated":642,"knowledge":643,"status":29,"reasoning":30,"reasoningControl":646,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":647,"openrouterUrl":648,"hf":-1,"company":649,"articles":655,"createdAt":683},"entity_01kyam49q1fvgvmfzzq6zh9173","Claude Opus 5","claude-opus-5","Anthropic","Claude","Opus 5","Opus 5 is a step change improvement for the Opus tier powering long-running agents while delivering improvements in coding and professional work.","https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fclaude-opus-5",1000000,128000,[16,17,18],[16],5,25,6.25,{"value":641,"precision":25},"2026-07-24",{"value":641,"precision":25},{"value":644,"precision":645},"2026-05-01","month",[32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fanthropic\u002Fclaude-opus-5\u002F","https:\u002F\u002Fopenrouter.ai\u002Fanthropic\u002Fclaude-opus-5",{"id":650,"name":628,"slug":651,"url":652,"bio":653,"githubOrg":654},"entity_01kyn5fq0re4frj5cwy4br9j9w","anthropic","https:\u002F\u002Fwww.anthropic.com","Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.","anthropics",[656,667,673],{"id":657,"name":658,"slug":659,"summary":660,"url":661,"kind":42,"platform":54,"author":662,"authorHandle":-1,"publisher":663,"publishedAt":664,"about":665,"writtenBy":666,"createdAt":-1},"entity_01kyk6yb70exs84dc1t042mk4b","How to Run a Gauntlet Loop","how-to-run-a-gauntlet-loop","The prompting method behind Claude of Duty. Give the agent a bar it can't talk its way around, let it split the work, and never let the builder grade itself.","https:\u002F\u002Fsomethingbig.ai\u002Fgauntlet-loop","Matt Shumer","Something Big Is Happening","2026-07-27",[],[],{"id":668,"name":669,"slug":670,"summary":631,"url":632,"kind":118,"platform":54,"author":628,"authorHandle":-1,"publisher":628,"publishedAt":641,"about":671,"writtenBy":672,"createdAt":-1},"entity_01kyam2heafvgvmfzn21fjm3za","Introducing Claude Opus 5","introducing-claude-opus-5",[],[],{"id":674,"name":675,"slug":676,"summary":677,"url":678,"kind":42,"platform":43,"author":679,"authorHandle":680,"publisher":46,"publishedAt":641,"about":681,"writtenBy":682,"createdAt":-1},"entity_01kyamm203f23s315waj08tmah","The new rules of context engineering for Claude 5 models","the-new-rules-of-context-engineering-for-claude-5-models","I’ve written previously about how to best prompt the newest generation of Claude 5 models and work with them iteratively to discover what you want to build.","https:\u002F\u002Fx.com\u002Ftrq212\u002Fstatus\u002F2080710971228918066","Thariq","trq212",[],[],1784915568,{"id":685,"name":686,"slug":687,"provider":688,"family":689,"variant":690,"description":691,"url":692,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":693,"outputLimit":409,"modalities":694,"modalitiesOut":695,"inputPricePerMTok":696,"outputPricePerMTok":697,"cacheReadPricePerMTok":698,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":699,"lastUpdated":701,"knowledge":702,"status":29,"reasoning":30,"reasoningControl":704,"toolCall":30,"attachment":30,"aiSdkId":705,"modelsDevUrl":706,"openrouterUrl":707,"hf":-1,"company":708,"articles":713,"createdAt":723},"entity_01ky39vzpbfanrdm0qdgeexe2z","Gemini 3.5 Flash-Lite","gemini-3-5-flash-lite","Google","Gemini","3.5 Flash-Lite","Our fastest, most cost-effective 3.5-class model, delivering 350 output tokens per second.",null,1048576,[16,17,132,133,18],[16],0.3,2.5,0.03,{"value":700,"precision":25},"2026-07-21",{"value":700,"precision":25},{"value":703,"precision":645},"2026-03-01",[32],"gemini-3.5-flash-lite","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fgoogle\u002Fgemini-3.5-flash-lite\u002F","https:\u002F\u002Fopenrouter.ai\u002Fgoogle\u002Fgemini-3.5-flash-lite",{"id":709,"name":710,"slug":711,"url":712,"bio":-1,"githubOrg":711},"entity_01kyn5fyfrecys2w4rpsvyfqn7","Google DeepMind","google-deepmind","https:\u002F\u002Fdeepmind.google",[714],{"id":715,"name":716,"slug":717,"summary":718,"url":719,"kind":118,"platform":54,"author":720,"authorHandle":-1,"publisher":688,"publishedAt":700,"about":721,"writtenBy":722,"createdAt":-1},"entity_01ky6887w6f24s834ath7ncqz0","Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber","introducing-gemini-3-6-flash-3-5-flash-lite-and-3-5-flash-cyber","We’re introducing new Gemini models, including Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber.","https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fmodels-and-research\u002Fgemini-models\u002Fgemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber\u002F","Tulsee Doshi",[],[],1784669929,{"id":725,"name":726,"slug":727,"provider":688,"family":689,"variant":728,"description":729,"url":692,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":693,"outputLimit":409,"modalities":730,"modalitiesOut":731,"inputPricePerMTok":732,"outputPricePerMTok":733,"cacheReadPricePerMTok":734,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":735,"lastUpdated":736,"knowledge":737,"status":29,"reasoning":30,"reasoningControl":738,"toolCall":30,"attachment":30,"aiSdkId":739,"modelsDevUrl":740,"openrouterUrl":741,"hf":-1,"company":742,"articles":743,"createdAt":747},"entity_01ky39vz5tfanrdm0fdg1k1vkn","Gemini 3.6 Flash","gemini-3-6-flash","3.6 Flash","Our workhorse model that delivers better coding, knowledge work, and multimodal performance.",[16,17,132,133,18],[16],1.5,7.5,0.15,{"value":700,"precision":25},{"value":700,"precision":25},{"value":703,"precision":645},[32],"gemini-3.6-flash","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fgoogle\u002Fgemini-3.6-flash\u002F","https:\u002F\u002Fopenrouter.ai\u002Fgoogle\u002Fgemini-3.6-flash",{"id":709,"name":710,"slug":711,"url":712,"bio":-1,"githubOrg":711},[744],{"id":715,"name":716,"slug":717,"summary":718,"url":719,"kind":118,"platform":54,"author":720,"authorHandle":-1,"publisher":688,"publishedAt":700,"about":745,"writtenBy":746,"createdAt":-1},[],[],1784669928,{"id":749,"name":750,"slug":751,"provider":752,"family":753,"variant":754,"description":755,"url":692,"openWeights":30,"license":220,"parameters":756,"contextWindow":693,"outputLimit":69,"modalities":757,"modalitiesOut":758,"inputPricePerMTok":77,"outputPricePerMTok":77,"cacheReadPricePerMTok":77,"cacheWritePricePerMTok":77,"announcedAt":-1,"releasedAt":759,"lastUpdated":760,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":761,"toolCall":30,"attachment":13,"aiSdkId":762,"modelsDevUrl":763,"openrouterUrl":764,"hf":765,"company":775,"articles":781,"createdAt":791},"entity_01ky39vxggfanrdkzqjzddm5wm","Laguna S 2.1","laguna-s-2-1","Poolside","Laguna","S-2.1","The most capable agentic coding model in its weight class by a wide margin.","118B total, ~8B active (MoE)",[16],[16],{"value":700,"precision":25},{"value":700,"precision":25},[228],"laguna-s-2.1","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fpoolside\u002Flaguna-s-2.1\u002F","https:\u002F\u002Fopenrouter.ai\u002Fpoolside\u002Flaguna-s-2.1",{"fullName":766,"url":767,"downloads":768,"likes":769,"license":770,"pipelineTag":192,"tags":771,"lastModified":774},"poolside\u002FLaguna-S-2.1","https:\u002F\u002Fhuggingface.co\u002Fpoolside\u002FLaguna-S-2.1",90327,962,"openmdw-1.1",[82,83,772,192,762,513,88,89,773,243,109],"laguna","license:openmdw-1.1","2026-07-27T11:25:04.000Z",{"id":776,"name":752,"slug":777,"url":778,"bio":779,"githubOrg":780},"entity_01kyn5g2zze4frj5dpycn0p8fk","poolside","https:\u002F\u002Fpoolside.ai","Poolside is a foundation model company bringing intelligence to everywhere work gets done. Our mission is to drive abundance for humanity by creating artificial general intelligence.","poolsideai",[782],{"id":783,"name":784,"slug":785,"summary":786,"url":787,"kind":118,"platform":54,"author":788,"authorHandle":-1,"publisher":752,"publishedAt":700,"about":789,"writtenBy":790,"createdAt":-1},"entity_01ky6c2w0rf24s83ck6yqg79p4","Introducing Laguna S 2.1","introducing-laguna-s-2-1","Today we're releasing Laguna S 2.1, a significant step forward in our development of models that pursue longer horizon work and make effective use of reasoning.","https:\u002F\u002Fpoolside.ai\u002Fblog\u002Fintroducing-laguna-s-2-1","Poolside team",[],[],1784669926,{"id":793,"name":794,"slug":795,"provider":796,"family":797,"variant":798,"description":799,"url":800,"openWeights":30,"license":794,"parameters":-1,"contextWindow":693,"outputLimit":315,"modalities":801,"modalitiesOut":802,"inputPricePerMTok":803,"outputPricePerMTok":804,"cacheReadPricePerMTok":696,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":805,"lastUpdated":807,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":808,"toolCall":30,"attachment":30,"aiSdkId":795,"modelsDevUrl":809,"openrouterUrl":810,"hf":811,"company":821,"articles":827,"createdAt":836},"entity_01kyjfghh8eh0v8dt1z7xfs2z4","Kimi K3","kimi-k3","Moonshot AI","Kimi","K3","Kimi K3 is a 2.8T-parameter model built on Kimi Delta Attention and Attention Residuals, with native vision capabilities and a 1-million-token context window — the world's first open 3T-class model.","https:\u002F\u002Fhuggingface.co\u002Fmoonshotai\u002FKimi-K3",[16,17,132],[16],3,15,{"value":806,"precision":25},"2026-07-16",{"value":806,"precision":25},[228,32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fmoonshotai\u002Fkimi-k3\u002F","https:\u002F\u002Fopenrouter.ai\u002Fmoonshotai\u002Fkimi-k3",{"fullName":812,"url":800,"downloads":813,"likes":814,"license":79,"pipelineTag":80,"tags":815,"lastModified":820},"moonshotai\u002FKimi-K3",1565484,10577,[82,83,816,817,818,88,80,89,107,243,819,109],"kimi_k3","feature-extraction","compressed-tensors","8-bit","2026-07-27T16:29:18.000Z",{"id":822,"name":796,"slug":823,"url":824,"bio":825,"githubOrg":826},"entity_01kyn5fyybecys2w550kwd1nfh","moonshot-ai","https:\u002F\u002Fwww.moonshot.ai","Welcome to Moonshot AI. Our mission is to seek the optimal conversion from energy to intelligence.","MoonshotAI",[828],{"id":829,"name":830,"slug":831,"summary":832,"url":833,"kind":118,"platform":54,"author":797,"authorHandle":-1,"publisher":797,"publishedAt":664,"about":834,"writtenBy":835,"createdAt":-1},"entity_01kyjf9j3zeh0v8ds8j505r64j","Kimi K3: Open Frontier Intelligence","kimi-k3-open-frontier-intelligence","Kimi K3 is the world's first open 3T-class model — frontier performance across coding, knowledge work, and reasoning, with native multimodality and 1M context.","https:\u002F\u002Fwww.kimi.com\u002Fblog\u002Fkimi-k3",[],[],1785179162,{"id":838,"name":589,"slug":839,"provider":588,"family":589,"variant":-1,"description":840,"url":841,"openWeights":30,"license":-1,"parameters":-1,"contextWindow":409,"outputLimit":409,"modalities":842,"modalitiesOut":843,"inputPricePerMTok":844,"outputPricePerMTok":845,"cacheReadPricePerMTok":846,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":847,"lastUpdated":849,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":850,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":851,"openrouterUrl":852,"hf":853,"company":860,"articles":861,"createdAt":872},"entity_01kyajqjxwesmb8qzjhcr1fkqa","inkling","Thinking Machines Lab's first open-weights model: a multimodal Mixture-of-Experts reasoner (975B total, 41B active) with controllable reasoning effort, fine-tunable on Tinker.","https:\u002F\u002Fthinkingmachines.ai\u002Fnews\u002Fintroducing-inkling\u002F",[16,17],[16],1.87,4.68,0.374,{"value":848,"precision":25},"2026-07-15",{"value":848,"precision":25},[228,32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fthinkingmachines\u002FInkling\u002F","https:\u002F\u002Fopenrouter.ai\u002Fthinkingmachines\u002Finkling",{"fullName":854,"url":855,"downloads":856,"likes":857,"license":191,"pipelineTag":80,"tags":858,"lastModified":859},"thinkingmachines\u002FInkling","https:\u002F\u002Fhuggingface.co\u002Fthinkingmachines\u002FInkling",84202,1726,[82,83,607,80,88,608,609,202,243,108,109],"2026-07-23T17:27:17.000Z",{"id":612,"name":588,"slug":613,"url":614,"bio":-1,"githubOrg":613},[862,865],{"id":617,"name":618,"slug":619,"summary":620,"url":592,"kind":42,"platform":54,"author":588,"authorHandle":-1,"publisher":588,"publishedAt":596,"about":863,"writtenBy":864,"createdAt":-1},[],[],{"id":866,"name":867,"slug":868,"summary":869,"url":841,"kind":42,"platform":54,"author":588,"authorHandle":-1,"publisher":588,"publishedAt":848,"about":870,"writtenBy":871,"createdAt":-1},"entity_01kyajeh43f6f93ar89zr8zsvg","Inkling: Our Open-Weights Model","inkling-our-open-weights-model","Our first open-weights model: multimodal, Mixture-of-Experts, with controllable reasoning effort. Available to fine-tune on Tinker.",[],[],1784914103,{"id":874,"name":875,"slug":876,"provider":877,"family":878,"variant":879,"description":880,"url":881,"openWeights":30,"license":180,"parameters":-1,"contextWindow":222,"outputLimit":-1,"modalities":882,"modalitiesOut":883,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":884,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":886,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":-1,"company":887,"articles":892,"createdAt":901},"entity_01ky944pz5e4a93wen7vyy88b8","Bonsai 27B","bonsai-27b","PrismML","Bonsai","27B","The first 27B-class model to run on a phone, based on Qwen3.6 27B: a multimodal flagship shipping in ternary (5.9 GB) and 1-bit (3.9 GB) forms with speculative decoding.","https:\u002F\u002Fhuggingface.co\u002Fcollections\u002Fprism-ml\u002Fbonsai-27b",[16,17],[16],{"value":885,"precision":25},"2026-07-14",[],{"id":888,"name":877,"slug":889,"url":890,"bio":891,"githubOrg":-1},"entity_01kyn5g3a4e4frj5dwb6gys6zs","prismml","https:\u002F\u002Fprismml.com\u002F","Large models can't fit on smartphones. Datacenters can't sustain them. PrismML is building ultra dense intelligence to solve both.",[893],{"id":894,"name":895,"slug":896,"summary":897,"url":898,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":877,"publishedAt":885,"about":899,"writtenBy":900,"createdAt":-1},"entity_01ky945c1be4a93wfdzfkw7c7q","Announcing Bonsai 27B: The First 27B-Class Model to Run on a Phone","announcing-bonsai-27b-the-first-27b-class-model-to-run-on-a-phone","Today, we're announcing Bonsai 27B, based on Qwen3.6 27B, the new multimodal flagship of the Bonsai family and the first model of its capability class to run on a phone.","https:\u002F\u002Fprismml.com\u002Fnews\u002Fbonsai-27b",[],[],1784865250,{"id":903,"name":904,"slug":905,"provider":906,"family":907,"variant":908,"description":909,"url":910,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":911,"outputLimit":634,"modalities":912,"modalitiesOut":913,"inputPricePerMTok":914,"outputPricePerMTok":915,"cacheReadPricePerMTok":916,"cacheWritePricePerMTok":917,"announcedAt":-1,"releasedAt":918,"lastUpdated":920,"knowledge":921,"status":29,"reasoning":30,"reasoningControl":923,"toolCall":30,"attachment":30,"aiSdkId":924,"modelsDevUrl":925,"openrouterUrl":926,"hf":-1,"company":927,"articles":931,"createdAt":958},"entity_01ky6a17n8f24s837acejbgyn1","GPT-5.6 Luna","gpt-5-6-luna","OpenAI","GPT-5","5.6 Luna","GPT-5.6 model optimized for cost-sensitive workloads.","https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fmodels\u002Fgpt-5.6-luna",1050000,[16,17,18],[16],0.2,1.2,0.02,0.25,{"value":919,"precision":25},"2026-07-09",{"value":919,"precision":25},{"value":922,"precision":25},"2026-02-16",[32],"gpt-5.6-luna","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fopenai\u002Fgpt-5.6-luna\u002F","https:\u002F\u002Fopenrouter.ai\u002Fopenai\u002Fgpt-5.6-luna",{"id":928,"name":906,"slug":929,"url":930,"bio":-1,"githubOrg":929},"entity_01kyn5g2gfe4frj5d9e0pvmtdf","openai","https:\u002F\u002Fopenai.com",[932,940,950],{"id":933,"name":934,"slug":935,"summary":936,"url":937,"kind":42,"platform":54,"author":906,"authorHandle":-1,"publisher":906,"publishedAt":596,"about":938,"writtenBy":939,"createdAt":-1},"entity_01kytd8znde8jsra6zrcv042j6","Advancing the price-performance frontier with GPT-5.6","advancing-the-price-performance-frontier-with-gpt-5-6","By making every layer more efficient, OpenAI is delivering stronger performance per dollar across more enterprise workloads.","https:\u002F\u002Fopenai.com\u002Findex\u002Fadvancing-the-price-performance-frontier-with-gpt-5-6\u002F",[],[],{"id":941,"name":942,"slug":943,"summary":944,"url":945,"kind":42,"platform":43,"author":946,"authorHandle":947,"publisher":46,"publishedAt":664,"about":948,"writtenBy":949,"createdAt":-1},"entity_01kyxet11rf6ebgh8dnfjcnnq0","Getting the most out of GPT-5.6: Sol, Terra, and Luna","getting-the-most-out-of-gpt-5-6-sol-terra-and-luna","Your Codex subscription now comes with 3 main models, Sol, Terra, and Luna, each independently trained and served, with reasoning dials.","https:\u002F\u002Fx.com\u002Fcerebras\u002Fstatus\u002F2081828128952095022","Cerebras","cerebras",[],[],{"id":951,"name":952,"slug":953,"summary":954,"url":955,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":906,"publishedAt":919,"about":956,"writtenBy":957,"createdAt":-1},"entity_01ky6a182sf24s837je9m950a1","GPT-5.6: Frontier intelligence that scales with your ambition","gpt-5-6","We’re launching the GPT‑5.6 family of models for general availability following our limited preview⁠: our new flagship, Sol, alongside Terra, a balanced model for everyday work, and Luna, our most cost-efficient model.","https:\u002F\u002Fopenai.com\u002Findex\u002Fgpt-5-6\u002F",[],[],1784770764,{"id":960,"name":961,"slug":962,"provider":906,"family":907,"variant":963,"description":964,"url":965,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":911,"outputLimit":634,"modalities":966,"modalitiesOut":967,"inputPricePerMTok":637,"outputPricePerMTok":968,"cacheReadPricePerMTok":22,"cacheWritePricePerMTok":639,"announcedAt":-1,"releasedAt":969,"lastUpdated":970,"knowledge":971,"status":29,"reasoning":30,"reasoningControl":972,"toolCall":30,"attachment":30,"aiSdkId":973,"modelsDevUrl":974,"openrouterUrl":975,"hf":-1,"company":976,"articles":977,"createdAt":998},"entity_01ky6a15nkf24s836rbzrtcrj9","GPT-5.6 Sol","gpt-5-6-sol","5.6 Sol","Frontier model for complex professional work.","https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fmodels\u002Fgpt-5.6-sol",[16,17,18],[16],30,{"value":919,"precision":25},{"value":919,"precision":25},{"value":922,"precision":25},[32],"gpt-5.6-sol","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fopenai\u002Fgpt-5.6-sol\u002F","https:\u002F\u002Fopenrouter.ai\u002Fopenai\u002Fgpt-5.6-sol",{"id":928,"name":906,"slug":929,"url":930,"bio":-1,"githubOrg":929},[978,981,984,995],{"id":933,"name":934,"slug":935,"summary":936,"url":937,"kind":42,"platform":54,"author":906,"authorHandle":-1,"publisher":906,"publishedAt":596,"about":979,"writtenBy":980,"createdAt":-1},[],[],{"id":941,"name":942,"slug":943,"summary":944,"url":945,"kind":42,"platform":43,"author":946,"authorHandle":947,"publisher":46,"publishedAt":664,"about":982,"writtenBy":983,"createdAt":-1},[],[],{"id":985,"name":986,"slug":987,"summary":988,"url":989,"kind":990,"platform":43,"author":991,"authorHandle":992,"publisher":46,"publishedAt":641,"about":993,"writtenBy":994,"createdAt":-1},"entity_01kyaxy91secrr44e185hbmdpz","Practical multi-agent orchestration in Codex","practical-multi-agent-orchestration-in-codex","GPT-5.6 Sol gets especially interesting when it has a team to work with.","https:\u002F\u002Fx.com\u002Fpvncher\u002Fstatus\u002F2080707291603407077","tutorial","eric provencher","pvncher",[],[],{"id":951,"name":952,"slug":953,"summary":954,"url":955,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":906,"publishedAt":919,"about":996,"writtenBy":997,"createdAt":-1},[],[],1784770762,{"id":1000,"name":1001,"slug":1002,"provider":906,"family":907,"variant":1003,"description":1004,"url":1005,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":911,"outputLimit":634,"modalities":1006,"modalitiesOut":1007,"inputPricePerMTok":20,"outputPricePerMTok":1008,"cacheReadPricePerMTok":914,"cacheWritePricePerMTok":697,"announcedAt":-1,"releasedAt":1009,"lastUpdated":1010,"knowledge":1011,"status":29,"reasoning":30,"reasoningControl":1012,"toolCall":30,"attachment":30,"aiSdkId":1013,"modelsDevUrl":1014,"openrouterUrl":1015,"hf":-1,"company":1016,"articles":1017,"createdAt":998},"entity_01ky6a165sf24s8377yqgpq1e2","GPT-5.6 Terra","gpt-5-6-terra","5.6 Terra","GPT-5.6 model that balances intelligence and cost.","https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fmodels\u002Fgpt-5.6-terra",[16,17,18],[16],12,{"value":919,"precision":25},{"value":919,"precision":25},{"value":922,"precision":25},[32],"gpt-5.6-terra","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fopenai\u002Fgpt-5.6-terra\u002F","https:\u002F\u002Fopenrouter.ai\u002Fopenai\u002Fgpt-5.6-terra",{"id":928,"name":906,"slug":929,"url":930,"bio":-1,"githubOrg":929},[1018,1021,1024],{"id":933,"name":934,"slug":935,"summary":936,"url":937,"kind":42,"platform":54,"author":906,"authorHandle":-1,"publisher":906,"publishedAt":596,"about":1019,"writtenBy":1020,"createdAt":-1},[],[],{"id":941,"name":942,"slug":943,"summary":944,"url":945,"kind":42,"platform":43,"author":946,"authorHandle":947,"publisher":46,"publishedAt":664,"about":1022,"writtenBy":1023,"createdAt":-1},[],[],{"id":951,"name":952,"slug":953,"summary":954,"url":955,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":906,"publishedAt":919,"about":1025,"writtenBy":1026,"createdAt":-1},[],[],{"id":1028,"name":1029,"slug":1030,"provider":8,"family":9,"variant":1031,"description":1032,"url":1033,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":14,"outputLimit":14,"modalities":1034,"modalitiesOut":1035,"inputPricePerMTok":20,"outputPricePerMTok":21,"cacheReadPricePerMTok":696,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1036,"lastUpdated":1038,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":1039,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":1040,"openrouterUrl":1041,"hf":-1,"company":1042,"articles":1047,"createdAt":1054},"entity_01ky94ge69fexsb0ht6ey9sr0s","Grok 4.5","grok-4-5","4.5","xAI's Grok 4.5 flagship for chat, coding, and agentic tool use, with lower hallucination risk.","https:\u002F\u002Fx.ai\u002Fnews\u002Fgrok-4-5",[16,17,18],[16],{"value":1037,"precision":25},"2026-07-08",{"value":1037,"precision":25},[32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fxai\u002Fgrok-4.5\u002F","https:\u002F\u002Fopenrouter.ai\u002Fx-ai\u002Fgrok-4.5",{"id":1043,"name":8,"slug":1044,"url":1045,"bio":-1,"githubOrg":1046},"entity_01kyn5g3wvecys2w58yt57x8pm","xai","https:\u002F\u002Fx.ai","xai-org",[1048],{"id":1049,"name":1050,"slug":1030,"summary":1051,"url":1033,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":8,"publishedAt":806,"about":1052,"writtenBy":1053,"createdAt":-1},"entity_01ky94hm21fexsb0jgjyysavm6","Introducing Grok 4.5","Today, we're launching Grok 4.5, SpaceXAI's smartest model built to excel at coding, agentic tasks, and knowledge work. It's our strongest model ever and was trained alongside Cursor.",[],[],1784865634,{"id":1056,"name":1057,"slug":1058,"provider":1059,"family":1060,"variant":1061,"description":1062,"url":1063,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":-1,"outputLimit":-1,"modalities":1064,"modalitiesOut":1065,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1066,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":1067,"toolCall":30,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":-1,"company":1068,"articles":1074,"createdAt":1075},"entity_01kzqbyckzfq9b4s5f838mg0yk","SWE-1.7","swe-1-7","Cognition","SWE","1.7","The most capable model Cognition has trained so far. It reaches frontier-level intelligence at a much lower cost, advancing the cost-performance Pareto curve.","https:\u002F\u002Fcognition.com\u002Fblog\u002Fswe-1-7",[16],[16],{"value":1037,"precision":25},[],{"id":1069,"name":1059,"slug":1070,"url":1071,"bio":1072,"githubOrg":1073},"entity_01kzmf2c2be8krkm8xm4dd2v37","cognition","https:\u002F\u002Fcognition.ai","Cognition operates Devin, the first autonomous software engineer. Devin plans, writes, tests, and ships production code inside your existing workflows.","cognitionai",[],1786416935,{"id":1077,"name":1078,"slug":1079,"provider":628,"family":629,"variant":1080,"description":1081,"url":1082,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":633,"outputLimit":634,"modalities":1083,"modalitiesOut":1084,"inputPricePerMTok":20,"outputPricePerMTok":1085,"cacheReadPricePerMTok":914,"cacheWritePricePerMTok":697,"announcedAt":-1,"releasedAt":1086,"lastUpdated":1088,"knowledge":1090,"status":29,"reasoning":30,"reasoningControl":1092,"toolCall":30,"attachment":30,"aiSdkId":-1,"modelsDevUrl":1093,"openrouterUrl":1094,"hf":-1,"company":1095,"articles":1096,"createdAt":1105},"entity_01kyzyeefdf6c8sdbvcrbqkys4","Claude Sonnet 5","claude-sonnet-5","Sonnet 5","Claude Sonnet 5 is built to be the most agentic Sonnet model yet. It can make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months ago, required larger and more expensive models.","https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fclaude-sonnet-5",[16,17,18],[16],10,{"value":1087,"precision":25},"2026-06-29",{"value":1089,"precision":25},"2026-06-30",{"value":1091,"precision":25},"2026-01-31",[228,32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fanthropic\u002Fclaude-sonnet-5\u002F","https:\u002F\u002Fopenrouter.ai\u002Fanthropic\u002Fclaude-sonnet-5",{"id":650,"name":628,"slug":651,"url":652,"bio":653,"githubOrg":654},[1097],{"id":1098,"name":1099,"slug":1100,"summary":1101,"url":1082,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":1102,"publishedAt":1089,"about":1103,"writtenBy":1104,"createdAt":-1},"entity_01kyzyf8t7f6c8sdchvgmmagsc","Introducing Claude Sonnet 5","introducing-claude-sonnet-5","Our most agentic Sonnet yet, with top-tier intelligence for coding and everyday professional work.","anthropic.com",[],[],1785631029,{"id":1107,"name":1108,"slug":1109,"provider":1110,"family":1111,"variant":1112,"description":1113,"url":1114,"openWeights":30,"license":281,"parameters":-1,"contextWindow":633,"outputLimit":315,"modalities":1115,"modalitiesOut":1116,"inputPricePerMTok":1117,"outputPricePerMTok":1118,"cacheReadPricePerMTok":1119,"cacheWritePricePerMTok":77,"announcedAt":-1,"releasedAt":1120,"lastUpdated":1122,"knowledge":-1,"status":29,"reasoning":30,"reasoningControl":1123,"toolCall":30,"attachment":13,"aiSdkId":1124,"modelsDevUrl":1125,"openrouterUrl":1126,"hf":1127,"company":1136,"articles":1142,"createdAt":1143},"entity_01ky4a2cjafexrjmcq4anjbkff","GLM-5.2","glm-5-2","Zhipu AI","GLM","5.2","We're introducing GLM-5.2, our latest flagship model for long-horizon tasks.","https:\u002F\u002Fhuggingface.co\u002Fzai-org\u002FGLM-5.2",[16],[16],1.4,4.4,0.26,{"value":1121,"precision":25},"2026-06-13",{"value":1121,"precision":25},[32],"glm-5.2","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fzhipuai\u002Fglm-5.2\u002F","https:\u002F\u002Fopenrouter.ai\u002Fz-ai\u002Fglm-5.2",{"fullName":1128,"url":1114,"downloads":1129,"likes":1130,"license":293,"pipelineTag":192,"tags":1131,"lastModified":1135},"zai-org\u002FGLM-5.2",2517575,4942,[82,83,1132,192,88,92,91,1133,1134,296,243,108,109],"glm_moe_dsa","arxiv:2602.15763","arxiv:2603.12201","2026-07-02T08:08:14.000Z",{"id":1137,"name":1110,"slug":1138,"url":1139,"bio":1140,"githubOrg":1141},"entity_01kyn5g46pe4frj5e9m1vy0qdh","zhipu-ai","https:\u002F\u002Fz.ai","Meet Z.ai, the AI assistant powered by GLM-5.2. Build websites, write code, handle long-horizon tasks, and get instant answers. Fast, smart, and reliable.","zai-org",[],1784703693,{"id":1145,"name":1146,"slug":1147,"provider":1148,"family":1148,"variant":1149,"description":1150,"url":1151,"openWeights":30,"license":281,"parameters":-1,"contextWindow":633,"outputLimit":1152,"modalities":1153,"modalitiesOut":1154,"inputPricePerMTok":1155,"outputPricePerMTok":1156,"cacheReadPricePerMTok":1157,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1158,"lastUpdated":1160,"knowledge":1161,"status":29,"reasoning":30,"reasoningControl":1163,"toolCall":30,"attachment":13,"aiSdkId":-1,"modelsDevUrl":1164,"openrouterUrl":1165,"hf":1166,"company":1175,"articles":1181,"createdAt":1193},"entity_01kywp2zgae4dr2vr8wkzjd3ca","DeepSeek V4 Flash","deepseek-v4-flash","DeepSeek","V4 Flash","DeepSeek-V4-Flash-0731 is the official release of DeepSeek-V4-Flash, superseding the preview version, with substantially enhanced agentic capabilities.","https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-V4-Flash-0731",384000,[16],[16],0.14,0.28,0.0028,{"value":1159,"precision":25},"2026-04-24",{"value":550,"precision":25},{"value":1162,"precision":645},"2025-05-01",[228,32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fdeepseek\u002Fdeepseek-v4-flash\u002F","https:\u002F\u002Fopenrouter.ai\u002Fdeepseek\u002Fdeepseek-v4-flash-0731",{"fullName":1167,"url":1151,"downloads":1168,"likes":1169,"license":293,"pipelineTag":192,"tags":1170,"lastModified":1174},"deepseek-ai\u002FDeepSeek-V4-Flash-0731",1048685,3227,[82,83,1171,192,88,1172,296,243,108,819,1173,244,109],"deepseek_v4","arxiv:2606.19348","fp8","2026-08-01T03:07:41.000Z",{"id":1176,"name":1148,"slug":1177,"url":1178,"bio":1179,"githubOrg":1180},"entity_01kywngn2se4dr2vpa6jd3ax4b","deepseek","https:\u002F\u002Fwww.deepseek.com\u002F","DeepSeek, unravel the mystery of AGI with curiosity. Answer the essential question with long-termism.","deepseek-ai",[1182],{"id":1183,"name":1184,"slug":1185,"summary":1186,"url":1187,"kind":1188,"platform":43,"author":1189,"authorHandle":1190,"publisher":46,"publishedAt":550,"about":1191,"writtenBy":1192,"createdAt":-1},"entity_01kywq4erfexyby29ded4bjday","We are making the updated DeepSeek V4-Flash 0731 free in Cline.","we-are-making-the-updated-deepseek-v4-flash-0731-free-in-cline","We are making the updated DeepSeek V4-Flash 0731 free in Cline.\n\nThis is the first flash model we've found performs at SOTA levels, and are excited for you to feel the new frontier.\n\n1. npm i -g cline\n2. Open \u002Fsettings > Cline provider\n3. Select deepseek-v4-flash","https:\u002F\u002Fx.com\u002Fcline\u002Fstatus\u002F2083249360662659079","post","Cline","cline",[],[],1785521602,{"id":1195,"name":1196,"slug":1197,"provider":1148,"family":1148,"variant":1198,"description":1199,"url":1200,"openWeights":30,"license":281,"parameters":1201,"contextWindow":633,"outputLimit":1152,"modalities":1202,"modalitiesOut":1203,"inputPricePerMTok":1204,"outputPricePerMTok":1205,"cacheReadPricePerMTok":1206,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1207,"lastUpdated":1208,"knowledge":1209,"status":29,"reasoning":30,"reasoningControl":1210,"toolCall":30,"attachment":13,"aiSdkId":-1,"modelsDevUrl":1211,"openrouterUrl":1212,"hf":1213,"company":-1,"articles":1219,"createdAt":1220},"entity_01kzvqhfq3fqcve1gscew13ycw","DeepSeek V4 Pro","deepseek-v4-pro","V4 Pro","DeepSeek-V4-Pro is a 1.6T-parameter Mixture-of-Experts model with 49B activated, supporting a one-million-token context. The 0813 revision is its GA release.","https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-V4-Pro","1.6T total, 49B active (MoE)",[16],[16],0.435,0.87,0.003625,{"value":1159,"precision":25},{"value":24,"precision":25},{"value":1162,"precision":645},[228,32],"https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fdeepseek\u002Fdeepseek-v4-pro\u002F","https:\u002F\u002Fopenrouter.ai\u002Fdeepseek\u002Fdeepseek-v4-pro-0813",{"fullName":1214,"url":1200,"downloads":1215,"likes":1216,"license":293,"pipelineTag":192,"tags":1217,"lastModified":1218},"deepseek-ai\u002FDeepSeek-V4-Pro",1411583,5408,[82,83,1171,192,88,1172,296,243,108,819,1173,109],"2026-06-22T12:12:50.000Z",[],1786563313,{"id":1222,"name":1223,"slug":1224,"provider":906,"family":907,"variant":1225,"description":1226,"url":692,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":911,"outputLimit":634,"modalities":1227,"modalitiesOut":1228,"inputPricePerMTok":637,"outputPricePerMTok":968,"cacheReadPricePerMTok":22,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1229,"lastUpdated":1231,"knowledge":1232,"status":29,"reasoning":30,"reasoningControl":1234,"toolCall":30,"attachment":30,"aiSdkId":1235,"modelsDevUrl":1236,"openrouterUrl":1237,"hf":-1,"company":1238,"articles":1239,"createdAt":1258},"entity_01ky68y5ghf24s834xgx83np63","GPT-5.5","gpt-5-5","5.5","A new class of intelligence for coding and professional work.",[16,17,18],[16],{"value":1230,"precision":25},"2026-04-23",{"value":1230,"precision":25},{"value":1233,"precision":25},"2025-12-01",[32],"gpt-5.5","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fopenai\u002Fgpt-5.5\u002F","https:\u002F\u002Fopenrouter.ai\u002Fopenai\u002Fgpt-5.5",{"id":928,"name":906,"slug":929,"url":930,"bio":-1,"githubOrg":929},[1240,1251],{"id":1241,"name":1242,"slug":1243,"summary":1244,"url":1245,"kind":42,"platform":43,"author":1246,"authorHandle":1247,"publisher":46,"publishedAt":1248,"about":1249,"writtenBy":1250,"createdAt":-1},"entity_01ky652cwefk1832njwmckzjgr","Random list of tips (ymmv) from trying to get gpt 5.5 to build a big thing over a day or two","random-list-of-tips-ymmv-from-trying-to-get-gpt-5-5-to-build-a-big-thi","","https:\u002F\u002Fx.com\u002Fgfodor\u002Fstatus\u002F2058931166405931452","gfodor.id","gfodor","2026-05-25",[],[],{"id":1252,"name":1253,"slug":1254,"summary":1226,"url":1255,"kind":118,"platform":54,"author":-1,"authorHandle":-1,"publisher":906,"publishedAt":1230,"about":1256,"writtenBy":1257,"createdAt":-1},"entity_01ky69t7mqf24s83635amb5f8s","Introducing GPT-5.5","introducing-gpt-5-5","https:\u002F\u002Fopenai.com\u002Findex\u002Fintroducing-gpt-5-5\u002F",[],[],1784769615,{"id":1260,"name":1261,"slug":1262,"provider":796,"family":797,"variant":1263,"description":1264,"url":1265,"openWeights":30,"license":1266,"parameters":1267,"contextWindow":222,"outputLimit":222,"modalities":1268,"modalitiesOut":1269,"inputPricePerMTok":1270,"outputPricePerMTok":1271,"cacheReadPricePerMTok":1272,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":1273,"lastUpdated":1275,"knowledge":1276,"status":29,"reasoning":30,"reasoningControl":1278,"toolCall":30,"attachment":30,"aiSdkId":1279,"modelsDevUrl":1280,"openrouterUrl":1281,"hf":1282,"company":1290,"articles":1291,"createdAt":1312},"entity_01kz6z0xr6eh3r946rqrbpks2t","Kimi K2.6","kimi-k2-6","K2.6","Kimi K2.6 is an open-source, native multimodal agentic model that advances practical capabilities in long-horizon coding, coding-driven design, proactive autonomous execution, and swarm-based task orchestration.","https:\u002F\u002Fhuggingface.co\u002Fmoonshotai\u002FKimi-K2.6","Modified MIT","1T",[16,17,132],[16],0.95,4,0.16,{"value":1274,"precision":25},"2026-04-21",{"value":1274,"precision":25},{"value":1277,"precision":645},"2025-01-01",[228],"kimi-k2.6","https:\u002F\u002Fmodels.dev\u002Fmodels\u002Fmoonshotai\u002Fkimi-k2.6\u002F","https:\u002F\u002Fopenrouter.ai\u002Fmoonshotai\u002Fkimi-k2.6",{"fullName":1283,"url":1265,"downloads":1284,"likes":1285,"license":79,"pipelineTag":80,"tags":1286,"lastModified":1289},"moonshotai\u002FKimi-K2.6",785559,1587,[82,83,1287,817,818,80,88,89,1288,107,243,109],"kimi_k25","arxiv:2602.02276","2026-05-19T09:01:54.000Z",{"id":822,"name":796,"slug":823,"url":824,"bio":825,"githubOrg":826},[1292,1302],{"id":1293,"name":1294,"slug":1295,"summary":1296,"url":1297,"kind":42,"platform":43,"author":1298,"authorHandle":1299,"publisher":46,"publishedAt":443,"about":1300,"writtenBy":1301,"createdAt":-1},"entity_01kz6ypgbqeh3r945vxxhpcgjs","How to Save Millions by Self-Hosting LLMs","how-to-save-millions-by-self-hosting-llms","At @cline we're on track to spend about $3M a year on inference for Kimi models. Everyone told us the same thing: it's open-weights, so just self-host and save.","https:\u002F\u002Fx.com\u002Farafatkatze\u002Fstatus\u002F2084668541270397223","Ara","arafatkatze",[],[],{"id":1303,"name":1304,"slug":1305,"summary":1306,"url":1307,"kind":42,"platform":54,"author":1308,"authorHandle":-1,"publisher":1309,"publishedAt":443,"about":1310,"writtenBy":1311,"createdAt":-1},"entity_01kz7zkk47en88gjs74evss0fr","How we built a software factory to drive Astro’s GitHub issue count to zero","how-we-built-a-software-factory-to-drive-astro-s-github-issue-count-to-zero","By replacing manual issue verification with isolated AI subagents running in GitHub Actions, the Astro maintainers reduced open issue count by 85%. This post explores the architecture behind automated bug reproduction, patch verification, and preview releases.","https:\u002F\u002Fblog.cloudflare.com\u002Fastro-issue-triage\u002F","Matthew Phillips","The Cloudflare Blog",[],[],1785866516,{"id":1314,"name":1315,"slug":1316,"provider":688,"family":689,"variant":1317,"description":1318,"url":692,"openWeights":13,"license":-1,"parameters":-1,"contextWindow":-1,"outputLimit":-1,"modalities":1319,"modalitiesOut":1320,"inputPricePerMTok":-1,"outputPricePerMTok":-1,"cacheReadPricePerMTok":-1,"cacheWritePricePerMTok":-1,"announcedAt":-1,"releasedAt":-1,"lastUpdated":-1,"knowledge":-1,"status":29,"reasoning":-1,"reasoningControl":1321,"toolCall":-1,"attachment":-1,"aiSdkId":-1,"modelsDevUrl":-1,"openrouterUrl":-1,"hf":-1,"company":1322,"articles":1323,"createdAt":723},"entity_01ky39w093fanrdm0s5rhqvsyw","Gemini 3.5 Flash-Cyber","gemini-3-5-flash-cyber","3.5 Flash-Cyber","Fine-tuned for finding and fixing cybersecurity vulnerabilities at a lower price per token than larger models.",[16],[],[],{"id":709,"name":710,"slug":711,"url":712,"bio":-1,"githubOrg":711},[1324],{"id":715,"name":716,"slug":717,"summary":718,"url":719,"kind":118,"platform":54,"author":720,"authorHandle":-1,"publisher":688,"publishedAt":700,"about":1325,"writtenBy":1326,"createdAt":-1},[],[]]