{"type":"data","nodes":[null,null,null,{"type":"data","data":[{"items":1,"total":469,"page":39,"pageSize":56,"hasMore":36,"freeTierActive":37,"bounds":470,"extraFilters":497,"scatterDetails":8,"error":8},[2,44,65,90,104,121,142,163,186,204,224,249,263,283,299,324,338,354,377,387,404,415,430,447],{"id":3,"category":4,"slug":5,"name":6,"description":7,"icon":8,"tags":9,"tier":23,"language":24,"aura_score":25,"aura_count":26,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":30,"brand_id":31,"brand_name":32,"brand_icon_light":33,"brand_icon_dark":33,"context_window":34,"capabilities":35,"input_cost":38,"output_cost":39,"image_cost":8,"audio_cost":8,"cache_read_cost":40,"model_type":15,"avg_latency_ms":41,"avg_tps":42,"avg_ttft_ms":43,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":5},"6f10337d-e76c-4a65-9e94-d6d0a7c96bdc","model","deepseek-v4.1-flash","DeepSeek V4.1 Flash","DeepSeek V4.1 Flash is a 552B-parameter multimodal Mixture-of-Experts (MoE) chat model from DeepSeek AI, activating only 8B parameters during prefill and 16B during decode thanks to its novel Causal Encoder-Decoder (CED) architecture with Compressed Sparse Attention 2 (CSA2). The model natively processes images and text, supports a 1M-token context window, and features continuously controllable reasoning effort (1–100) for trading inference cost against accuracy. It achieves 90.9 on GPQA Diamond, 90.6 on Terminal-Bench 2.1, and 74.2 on DeepSWE v1.1, demonstrating frontier-level agentic and coding performance. The model is released under the MIT License.",null,[10,11,12,13,14,15,16,17,18,19,20,21,22],"deepseek","v4.1","flash","552b","moe","chat","llm","multimodal","vision","mit-license","long-context","reasoning","agentic","free","en",30,36,"d03efaea-a33e-42ff-8dde-c3c2cb0ea560","Anonymous",0,"2026-09-10T13:42:49.866333Z","c5ac869a-1426-46cd-8f8d-2727c9d2faca","Deepseek","https://cdn.melious.ai/icons/brands/deepseek.svg",1000000,{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},true,false,0.2,1,0.01,455,270,396,{"id":45,"category":4,"slug":46,"name":47,"description":48,"icon":8,"tags":49,"tier":23,"language":24,"aura_score":56,"aura_count":25,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":57,"brand_id":31,"brand_name":32,"brand_icon_light":33,"brand_icon_dark":33,"context_window":34,"capabilities":58,"input_cost":59,"output_cost":60,"image_cost":8,"audio_cost":8,"cache_read_cost":61,"model_type":15,"avg_latency_ms":62,"avg_tps":63,"avg_ttft_ms":64,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":46},"031eb497-31d2-45e1-ac82-050e5335daab","deepseek-v4-flash-0731","DeepSeek V4 Flash 0731","DeepSeek V4 Flash 0731 is the official release of DeepSeek V4 Flash, superseding the preview version. It is a 304B-parameter Mixture-of-Experts model (256 routed experts plus one shared expert, 6 active per token) with an attached DSpark speculative-decoding module and a 1M-token context window. Deliberation is controlled through three reasoning effort levels — low, high and max — with up to 384K output tokens recommended at the high and max levels. On published agentic benchmarks it reaches 82.7 on Terminal Bench 2.1, 70.3 on Toolathlon-Verified and 54.4 on DeepSWE, outperforming DeepSeek V4 Pro (Preview) despite a far smaller activated parameter count. Text-only, with function calling and structured output, released under the MIT License.",[10,50,12,51,52,14,15,16,19,20,53,22,54,21,55],"v4","0731","304b","1m-context","coding","function-calling",24,"2026-07-31T15:19:39.396195Z",{"vision":37,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.1,0.25,0.02,725,184,322,{"id":66,"category":4,"slug":67,"name":68,"description":69,"icon":8,"tags":70,"tier":23,"language":24,"aura_score":78,"aura_count":79,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":80,"brand_id":81,"brand_name":82,"brand_icon_light":83,"brand_icon_dark":84,"context_window":34,"capabilities":85,"input_cost":59,"output_cost":86,"image_cost":8,"audio_cost":8,"cache_read_cost":61,"model_type":15,"avg_latency_ms":87,"avg_tps":88,"avg_ttft_ms":89,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":67},"753254ca-9bc8-4e8f-9968-923be5b25e31","glm-5.3-flash","GLM 5.3 Flash","ZAI GLM 5.3 Flash is a 320B parameter Mixture-of-Experts (MoE) language model with 18B active parameters, and the first natively multimodal model in the GLM-5 series. It introduces a hybrid architecture combining sparse and linear attention to reduce long-context serving costs, along with Manifold-Constrained Hyper-Connections (mHC) for improved scaling efficiency, trained on a 30T-token multimodal corpus. The model supports configurable reasoning effort levels (low, high, max) with thinking enabled by default, and features a context window of up to 1 million tokens. It outperforms GLM-5.2 across benchmarks and real-world workloads at one-tenth the price, while approaching Claude Opus 4.8 on coding and agentic benchmarks. Released under the MIT license.",[71,72,73,12,14,74,75,17,15,16,76,77,53,19,22,18],"zai","glm","5.3","320b","18b-active","hybrid-reasoning","tool-calling",18,28,"2026-08-30T22:41:22.244058Z","93989f12-5c7e-4f91-b0d3-c6b60f2de452","ZAI","https://cdn.melious.ai/icons/brands/zai_light.svg","https://cdn.melious.ai/icons/brands/zai_dark.svg",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.4,264,199,404,{"id":91,"category":4,"slug":92,"name":93,"description":94,"icon":8,"tags":95,"tier":23,"language":24,"aura_score":96,"aura_count":97,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":98,"brand_id":81,"brand_name":82,"brand_icon_light":83,"brand_icon_dark":84,"context_window":34,"capabilities":99,"input_cost":39,"output_cost":100,"image_cost":8,"audio_cost":8,"cache_read_cost":38,"model_type":15,"avg_latency_ms":101,"avg_tps":102,"avg_ttft_ms":103,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":92},"413e5b6f-db57-434a-9b7b-837eb15a8a8a","glm-5.3","GLM 5.3","ZAI GLM 5.3 is a 744B parameter Mixture-of-Experts (MoE) language model built on the GLM-MoE DSA architecture, using the same base model as GLM-5.2 with significant post-training improvements focused on complex coding and long-horizon agentic tasks. It features a 1-million token context window and configurable reasoning effort levels (low, high, max), with thinking enabled by default. The model achieves strong benchmark results including 88.2 on Terminal Bench 2.1, 66.9 on DeepSWE (v1.1), 62.5 on HLE with Tools, and 73.0 on Toolathlon Verified. It is released under a custom license.",[71,72,73,14,15,16,76,77,53,22,54],12,14,"2026-08-28T15:37:26.665777Z",{"vision":37,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},3,1049,94,666,{"id":105,"category":4,"slug":106,"name":107,"description":108,"icon":8,"tags":109,"tier":23,"language":24,"aura_score":111,"aura_count":112,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":113,"brand_id":31,"brand_name":32,"brand_icon_light":33,"brand_icon_dark":33,"context_window":34,"capabilities":114,"input_cost":115,"output_cost":116,"image_cost":8,"audio_cost":8,"cache_read_cost":117,"model_type":15,"avg_latency_ms":118,"avg_tps":119,"avg_ttft_ms":120,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":106},"36cc8716-50e6-45be-b576-f457a8350334","deepseek-v4.1-flash-1017","DeepSeek V4.1 Flash 1017","DeepSeek V4.1 Flash 1017 is a minor update to DeepSeek V4.1 Flash, a 552B-parameter multimodal Mixture-of-Experts (MoE) chat model built on the Causal Encoder-Decoder (CED) architecture with compressed KV cache for efficient long-context serving. The update delivers substantial improvements across reasoning, coding, agentic tasks, and visual benchmarks while reducing task execution time and output length, advancing the performance–cost Pareto frontier across multiple agent harnesses. It achieves 42.6 on HLE without tools, 42.3 on Terminal-Bench 4.0, 77.0 on DeepSWE v1.1, and 72.1 on Apex Bench v1.1. The model natively processes images and text, supports a 1M-token context window with up to 256K output tokens, and features thinking mode with numeric reasoning effort control. Available under the MIT license.",[10,11,12,110,13,14,15,16,17,18,21,22,20,19],"1017",8,20,"2026-10-19T11:09:23.147731Z",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.3,1.2,0.03,426,153,582,{"id":122,"category":4,"slug":123,"name":124,"description":125,"icon":8,"tags":126,"tier":23,"language":24,"aura_score":111,"aura_count":132,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":133,"brand_id":134,"brand_name":135,"brand_icon_light":136,"brand_icon_dark":136,"context_window":137,"capabilities":138,"input_cost":59,"output_cost":115,"image_cost":8,"audio_cost":8,"cache_read_cost":61,"model_type":15,"avg_latency_ms":139,"avg_tps":140,"avg_ttft_ms":141,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":123},"d50d214d-482d-4366-8dcb-e91fabbc1afd","gemma-4-31b","Gemma 4 31B","Google Gemma 4 31B is a 31B parameter dense multimodal language model with a 256K context window. It processes text, images, and video inputs and generates text output, featuring a configurable thinking mode for step‑by‑step reasoning. The model achieves 85.2% on MMLU Pro, 80.0% on LiveCodeBench v6, and 88.4% on MMMLU, demonstrating strong performance across reasoning and multimodal benchmarks. Available under the Apache 2.0 license.",[127,128,129,17,18,130,15,76,55,131,16],"google","gemma-4","31b","video","apache-2.0",10,"2026-04-14T18:35:29.619976Z","cc4a3d6e-7e99-4f28-9a1c-505a2ac39e0c","Google","https://cdn.melious.ai/icons/brands/google.svg",256000,{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},310,44,1142,{"id":143,"category":4,"slug":144,"name":145,"description":146,"icon":8,"tags":147,"tier":23,"language":24,"aura_score":148,"aura_count":149,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":151,"brand_id":152,"brand_name":153,"brand_icon_light":154,"brand_icon_dark":155,"context_window":34,"capabilities":156,"input_cost":157,"output_cost":158,"image_cost":8,"audio_cost":8,"cache_read_cost":159,"model_type":15,"avg_latency_ms":160,"avg_tps":161,"avg_ttft_ms":162,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":144},"cb0fe7cf-ddf6-4f68-b11a-9f5478100f37","kimi-k3","Kimi K3","Kimi K3 is an open-weight, native multimodal agentic model and our most capable model to date. It is a 2.8T-parameter model built on Kimi Delta Attention (KDA) and Attention Residuals (AttnRes), with native vision capabilities and a 1-million-token context window. It is the world's first open 3T-class model, designed for frontier intelligence across long-horizon coding, knowledge work, and reasoning.",[],5,11,"Melious","2026-07-27T15:42:58.418089Z","b724f389-794d-4495-9194-13210510019e","Moonshot","https://cdn.melious.ai/icons/brands/moonshot_light.svg","https://cdn.melious.ai/icons/brands/moonshot_dark.svg",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},2.85,14.25,0.75,29206,62,1438,{"id":164,"category":4,"slug":165,"name":166,"description":167,"icon":8,"tags":168,"tier":23,"language":24,"aura_score":176,"aura_count":176,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":177,"brand_id":178,"brand_name":179,"brand_icon_light":180,"brand_icon_dark":180,"context_window":181,"capabilities":182,"input_cost":86,"output_cost":183,"image_cost":8,"audio_cost":8,"cache_read_cost":59,"model_type":15,"avg_latency_ms":184,"avg_tps":185,"avg_ttft_ms":119,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":165},"96b7df44-471c-49b1-aac9-861397f4c74a","qwen3.8-27b","Qwen3.8 27B","Qwen3.8 27B is a 27B-parameter dense vision-language model built on the Qwen3.5 gated-delta architecture with a vision encoder, supporting a native 262K context window extensible to 1M tokens. It operates in a default thinking mode that can be disabled per request, with adjustable reasoning depth via reasoning_effort (xhigh, medium, low) and preserved thinking context across turns. The model achieves 89.2 on GPQA Diamond, 90.3 on LiveCodeBench v6, 61.7 on SWE-bench Pro, and 84.3 on OSWorld-Verified, demonstrating strong coding, agentic, and multimodal capabilities including native image and video understanding. It is released under the Apache 2.0 license.",[169,170,171,172,15,17,173,130,174,22,175,131,16],"qwen","qwen3.8","27b","dense","vision-language","thinking","computer-use",4,"2026-08-14T16:09:35.085758Z","73399f6c-c2ad-419a-a846-12ea92cb6c01","Qwen","https://cdn.melious.ai/icons/brands/qwen.svg",262000,{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},2.4,104,192,{"id":187,"category":4,"slug":188,"name":189,"description":190,"icon":8,"tags":191,"tier":23,"language":24,"aura_score":176,"aura_count":176,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":200,"brand_id":178,"brand_name":179,"brand_icon_light":180,"brand_icon_dark":180,"context_window":201,"capabilities":202,"input_cost":40,"output_cost":40,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":199,"avg_latency_ms":203,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":188},"e86194a8-9a2b-42ff-a946-c88d7911a36d","qwen3-embedding-8b","Qwen3 Embedding 8B","Qwen3 Embedding 8B is a dense retrieval embedding model with 8 billion parameters, optimized for semantic search, text similarity, and feature extraction. Trained on diverse multilingual data providing strong cross-lingual retrieval capabilities. Supports 262K context for embedding long documents and extensive text passages. Excels at document retrieval, semantic search, clustering, and recommendation systems. Compatible with standard embedding frameworks and optimized for production deployment with efficient inference.",[169,192,193,194,195,196,197,198,199],"embedding","8b","retrieval","semantic-search","multilingual","262k-context","text-similarity","embeddings","2025-12-16T15:32:07.832850Z",32000,{"vision":37,"function_calling":37,"streaming":36,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},71,{"id":205,"category":4,"slug":206,"name":207,"description":208,"icon":8,"tags":209,"tier":23,"language":24,"aura_score":100,"aura_count":100,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":218,"brand_id":152,"brand_name":153,"brand_icon_light":154,"brand_icon_dark":155,"context_window":137,"capabilities":219,"input_cost":220,"output_cost":221,"image_cost":8,"audio_cost":8,"cache_read_cost":38,"model_type":15,"avg_latency_ms":222,"avg_tps":119,"avg_ttft_ms":223,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":206},"23cb751b-fa48-4df9-a15e-6351aa1397fb","kimi-k2.6","Kimi K2.6","Moonshot Kimi K2.6 is a 1 trillion parameter Mixture-of-Experts chat model with 32 B activated parameters and a 256K context window. It combines a vision encoder (MoonViT) and supports image, video, and text inputs, enabling multimodal agentic interactions with tool calling and structured JSON output. The model achieves strong benchmark scores such as 91.7% on LiveCodeBench, 44.4% on HLE‑Full (reasoning) and 83.0% on MMMU‑Pro (vision) while offering a thinking mode for deep reasoning and an instant mode for fast responses. Available under an \"Other\" license.",[210,211,212,213,214,14,215,216,17,18,130,22,77,174,15,16,217],"moonshot","kimi","k2.6","1t","trillion","32b-active","256k-context","other-license","2026-04-23T08:52:44.838170Z",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.8,3.55,161,568,{"id":225,"category":4,"slug":226,"name":227,"description":228,"icon":8,"tags":229,"tier":23,"language":24,"aura_score":236,"aura_count":176,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":237,"brand_id":238,"brand_name":239,"brand_icon_light":240,"brand_icon_dark":241,"context_window":242,"capabilities":243,"input_cost":244,"output_cost":245,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":246,"avg_tps":247,"avg_ttft_ms":248,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":226},"af6df590-c9ef-429d-a866-e977bcae1b0f","gpt-oss-120b","GPT-OSS 120B","GPT-OSS 120B is a powerful 117B parameter Mixture-of-Experts reasoning model with 5.1B active parameters, released under Apache 2.0. Features configurable reasoning effort (low/medium/high), full chain-of-thought visibility, and runs on a single 80GB GPU thanks to MXFP4 quantization. Native support for function calling, web browsing, Python code execution, and structured outputs. Designed for agentic tasks and complex reasoning with production-grade performance. Fully customizable for specialized use cases on single H100/MI300X.",[230,231,232,14,233,21,55,22,131,234,235,16],"openai","gpt-oss","120b","mixture-of-experts","open-source","chain-of-thought",2,"2025-12-16T15:31:52.117560Z","290668c1-0d15-4863-b1c5-c74a4fb5a0a2","OpenAI","https://cdn.melious.ai/icons/brands/openai_light.svg","https://cdn.melious.ai/icons/brands/openai_dark.svg",128000,{"vision":37,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.09,0.45,289,99,172,{"id":250,"category":4,"slug":251,"name":252,"description":253,"icon":8,"tags":254,"tier":23,"language":24,"aura_score":236,"aura_count":256,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":257,"brand_id":178,"brand_name":179,"brand_icon_light":180,"brand_icon_dark":180,"context_window":181,"capabilities":258,"input_cost":245,"output_cost":259,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":260,"avg_tps":261,"avg_ttft_ms":262,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":36,"model_id":251},"21430683-db06-4465-a557-1a26db47d4b3","qwen3.6-27b","Qwen 3.6 27B","Qwen 3.6 27B is a 27B‑parameter causal language model with a vision encoder, built on the Qwen3.5 gated‑delta architecture. It supports a native context window of 262,144 tokens (extendable to over 1 million) and operates in a default thinking mode that can be disabled. The model achieves strong benchmark results such as 86.2% on MMLU‑Pro, 82.9% on MMMU, and 77.2% on SWE‑bench Verified, demonstrating strong coding and multimodal reasoning capabilities. It is released under the Apache 2.0 license.",[169,255,171,15,17,173,174,131,16],"qwen3.6",6,"2026-04-28T11:55:16.156498Z",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},3.15,180,56,913,{"id":264,"category":4,"slug":265,"name":266,"description":267,"icon":8,"tags":268,"tier":23,"language":24,"aura_score":236,"aura_count":236,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":280,"brand_id":238,"brand_name":239,"brand_icon_light":240,"brand_icon_dark":241,"context_window":8,"capabilities":281,"input_cost":29,"output_cost":29,"image_cost":8,"audio_cost":282,"cache_read_cost":8,"model_type":277,"avg_latency_ms":8,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":265},"50287dc8-e020-4c1d-86d5-31a6c76c4f87","whisper-large-v3-turbo","Whisper Large V3 Turbo","OpenAI Whisper Large V3 Turbo is an optimized variant of Whisper V3 with significantly faster inference while maintaining high accuracy across 99 languages. Features architectural optimizations for reduced latency including faster encoder-decoder inference and efficient attention mechanisms. Delivers near-V3 accuracy with 2-3x speed improvement, ideal for real-time transcription applications, live subtitling, and high-throughput ASR workloads. Supports full multilingual capabilities, timestamps, and speech translation to English. Perfect for production deployments requiring both quality and speed.",[230,269,270,271,272,273,274,275,276,277,278,279],"whisper","v3-turbo","large","fast","speech-recognition","transcription","99-languages","real-time","audio","optimized","asr","2025-12-16T15:32:16.662043Z",{"vision":37,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":36,"computer_use":37,"reasoning":37},0.001,{"id":284,"category":4,"slug":285,"name":286,"description":287,"icon":288,"tags":289,"tier":23,"language":24,"aura_score":39,"aura_count":39,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":293,"brand_id":294,"brand_name":295,"brand_icon_light":8,"brand_icon_dark":8,"context_window":296,"capabilities":297,"input_cost":117,"output_cost":29,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":290,"avg_latency_ms":298,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":285},"1daf5c0b-dbd4-4bba-84d8-b4bf35624439","clef-flash","Clef Flash","Cloudflare's Clef Flash decision model (9B), the fast and low-cost member of the Clef family. Typed decisions with probabilities: choice, score and yes/no questions about a text, through the System One API (POST /v1/systemone). Billed per input token; no output tokens.","https://cdn.melious.ai/icons/brands/cloudflare.svg",[290,291,292],"decision","system-one","classification","2026-10-17T08:17:34.924705Z","5b0bfe34-c2c5-4199-becb-605f0d63ebbc","Cloudflare",16384,{"vision":37,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},142,{"id":300,"category":4,"slug":301,"name":302,"description":303,"icon":8,"tags":304,"tier":23,"language":24,"aura_score":39,"aura_count":39,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":315,"brand_id":316,"brand_name":317,"brand_icon_light":318,"brand_icon_dark":319,"context_window":8,"capabilities":320,"input_cost":29,"output_cost":29,"image_cost":321,"audio_cost":8,"cache_read_cost":8,"model_type":322,"avg_latency_ms":8,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":323,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":301},"c6e59702-0653-4e55-8fbe-aadfba306ee2","flux-2-klein-9b","FLUX.2 [klein] 9B","Black Forest Labs FLUX.2 [klein] 9B is a balanced image generation model offering excellent quality-to-speed ratio. With 9 billion parameters, it provides better detail and composition than the 4B variant while remaining faster than full-size models. Ideal for production workloads requiring a balance between quality, speed, and cost. Supports both text-to-image and image-to-image generation.",[305,306,307,308,309,310,311,312,313,314],"black-forest-labs","flux","flux2","klein","9b","balanced","text-to-image","image-generation","production","quality","2026-02-03T14:28:07.059555Z","2bade5cc-aec1-4b8d-a7e1-be382ae9832d","Black Forest Labs","https://cdn.melious.ai/icons/brands/bfl_light.svg","https://cdn.melious.ai/icons/brands/bfl_dark.svg",{"vision":36,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},0.009,"image",1584,{"id":325,"category":4,"slug":326,"name":327,"description":328,"icon":8,"tags":329,"tier":23,"language":24,"aura_score":39,"aura_count":39,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":330,"brand_id":331,"brand_name":332,"brand_icon_light":333,"brand_icon_dark":333,"context_window":334,"capabilities":335,"input_cost":117,"output_cost":29,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":199,"avg_latency_ms":336,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":337,"free_tier_eligible":37,"model_id":326},"bdf41426-7be9-476e-9f4d-fe1d10230d68","multilingual-e5-large-instruct","Multilingual E5 Large Instruct","intfloat Multilingual E5 Large Instruct is an instruction-tuned multilingual embedding model combining strong cross-lingual capabilities with instruction-following for guided retrieval. Supports 100+ languages with natural language instructions to customize embedding behavior. Features enhanced zero-shot retrieval performance through instruction-based query understanding. Ideal for complex multilingual search scenarios, domain-specific retrieval tasks, and applications requiring adaptive semantic understanding across languages.",[],"2026-01-07T14:25:51.875523Z","e6cde81c-104a-4dad-9e30-02bd3b9e9b98","intfloat","https://cdn.melious.ai/icons/brands/intfloat.svg",512,{"vision":37,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},141,352,{"id":339,"category":4,"slug":340,"name":341,"description":342,"icon":8,"tags":343,"tier":23,"language":24,"aura_score":39,"aura_count":39,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":348,"brand_id":178,"brand_name":179,"brand_icon_light":180,"brand_icon_dark":180,"context_window":137,"capabilities":349,"input_cost":60,"output_cost":350,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":351,"avg_tps":352,"avg_ttft_ms":353,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":340},"3a2e67bf-4283-4806-8f8f-1e5a2f10b309","qwen3.6-35b-a3b","Qwen 3.6 35B A3B","Qwen 3.6 35B A3B is a 35 B parameter mixture‑of‑experts vision‑language model with a gated delta network architecture and a vision encoder. It supports a native context window of 262 K tokens (extendable to over 1 M tokens) and can be used for image‑text‑to‑text tasks. The model achieves strong performance on benchmarks, scoring 85.2 % on MMLU‑Pro, 81.4 % on MMMU, and 85.3 % on RealWorldQA. It is released under the Apache 2.0 license.",[169,255,344,345,14,173,17,346,15,77,347,131,16,197],"35b","a3b","image-text-to-text","streaming","2026-05-26T13:53:24.528961Z",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},1.5,131,182,463,{"id":355,"category":4,"slug":356,"name":357,"description":358,"icon":8,"tags":359,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":367,"brand_id":368,"brand_name":369,"brand_icon_light":370,"brand_icon_dark":370,"context_window":371,"capabilities":372,"input_cost":373,"output_cost":373,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":374,"avg_tps":375,"avg_ttft_ms":376,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":356},"c8422732-1827-467d-89de-865a221f056b","llama-3.3-70b-instruct","Llama 3.3 70B Instruct","Meta Llama 3.3 70B Instruct is a multilingual instruction-tuned model optimized for dialogue. Trained on ~15 trillion tokens with cutoff December 2023, it outperforms many open-source and closed models. Major improvements include 92.1% on IFEval (steerability), 88.4% on HumanEval (code), 77.0% on MATH, and 91.1% on MGSM (multilingual). Features 128K context, Grouped-Query Attention, and supports 8 languages including English, German, French, Spanish, Italian, Portuguese, Hindi, and Thai. Trained on 7M GPU hours with 100% renewable energy.",[360,361,362,196,363,364,21,365,366,16],"meta","llama","70b","code","math","128k-context","instruction-following","2025-12-16T15:31:56.545449Z","0588b92b-b0a9-4633-93c9-70ec1f33d970","Meta","https://cdn.melious.ai/icons/brands/meta.svg",131072,{"vision":37,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":37},0.65,7402,47,258,{"id":378,"category":4,"slug":379,"name":380,"description":381,"icon":8,"tags":382,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":330,"brand_id":331,"brand_name":332,"brand_icon_light":333,"brand_icon_dark":333,"context_window":383,"capabilities":384,"input_cost":117,"output_cost":29,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":199,"avg_latency_ms":385,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":386,"free_tier_eligible":37,"model_id":379},"3f75133c-c182-4801-b0ae-d6801b56ed13","multilingual-e5-large","Multilingual E5 Large","intfloat Multilingual E5 Large is a powerful multilingual dense retrieval embedding model supporting 100+ languages with strong cross-lingual capabilities. Features 1024-dimensional embeddings optimized for semantic search, document retrieval, and text similarity across diverse language families. Pre-trained on large-scale multilingual data with contrastive learning for robust cross-lingual transfer. Ideal for international search systems, multilingual document retrieval, and global content recommendation platforms requiring high-quality semantic understanding.",[],514,{"vision":37,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},132,447,{"id":388,"category":4,"slug":389,"name":390,"description":391,"icon":8,"tags":392,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":397,"brand_id":238,"brand_name":239,"brand_icon_light":240,"brand_icon_dark":241,"context_window":242,"capabilities":398,"input_cost":399,"output_cost":400,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":401,"avg_tps":402,"avg_ttft_ms":403,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":389},"8970e80d-534c-4c65-9e75-df6270ea828c","gpt-oss-20b","GPT-OSS 20B","GPT-OSS 20B is a compact 21B parameter Mixture-of-Experts model with 3.6B active parameters, designed for lower latency and local deployment. Runs within 16GB memory with configurable reasoning effort, full chain-of-thought access, and native agentic capabilities including function calling and structured outputs. Released under Apache 2.0 license, ideal for specialized fine-tuning on consumer hardware. Companion model to GPT-OSS 120B optimized for speed while maintaining strong reasoning capabilities.",[230,231,393,14,233,21,394,395,55,131,396,16],"20b","local-deployment","low-latency","efficient","2025-12-16T15:31:52.774400Z",{"vision":37,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.05,0.17,110,61,209,{"id":405,"category":4,"slug":406,"name":407,"description":408,"icon":8,"tags":409,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":412,"brand_id":238,"brand_name":239,"brand_icon_light":240,"brand_icon_dark":241,"context_window":8,"capabilities":413,"input_cost":29,"output_cost":29,"image_cost":8,"audio_cost":414,"cache_read_cost":8,"model_type":277,"avg_latency_ms":8,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":406},"be3cb81b-df0b-493a-9ad0-980a3a6c60d3","whisper-large-v3","Whisper Large V3","OpenAI Whisper Large V3 is a state-of-the-art automatic speech recognition model with 1550M parameters supporting 99 languages. Achieves 10-20% WER reduction compared to V2, trained on 1M hours weakly labeled + 4M hours pseudo-labeled audio. Features 128 Mel frequency bins (increased from 80), improved robustness to accents and background noise, and new Cantonese language support. Supports speech transcription and speech-to-English translation with sentence and word-level timestamps. Optimized with torch.compile for 4.5x speedup. Ideal for accessibility tools, multilingual transcription, and enterprise ASR applications.",[230,269,410,271,411,273,274,275,196,277,279],"v3","1550m","2025-12-16T15:32:16.033947Z",{"vision":37,"function_calling":37,"streaming":37,"structured_output":37,"supports_audio":36,"computer_use":37,"reasoning":37},0.003,{"id":416,"category":4,"slug":417,"name":418,"description":419,"icon":8,"tags":420,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":423,"brand_id":424,"brand_name":425,"brand_icon_light":426,"brand_icon_dark":426,"context_window":371,"capabilities":427,"input_cost":59,"output_cost":115,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":15,"avg_latency_ms":428,"avg_tps":375,"avg_ttft_ms":429,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":417},"dc05968f-8195-42aa-bdc5-6b2ccf9f7e16","mistral-small-3.2-24b-instruct","Mistral Small 3.2 24B Instruct","Mistral Small 3.2 24B Instruct is a multimodal instruction-tuned model supporting both vision and text with 24B parameters and 128K context. Major improvements over 3.1 include better instruction following (84.78%), 2x reduction in repetition errors, and robust function calling. Achieves 65.33% on Wildbench v2, 43.1% on Arena Hard v2, 92.90% on HumanEval Pass@5. Vision benchmarks: 87.4% ChartQA, 94.86% DocVQA, 62.50% MMMU. Supports up to 10 images per prompt with integrated vision-based function calling.",[421,422,17,18,55,366,365,363,16],"mistral","24b","2025-12-16T15:31:57.802722Z","144e8e0d-941f-4e8c-93d9-56cf7edb5a67","Mistral","https://cdn.melious.ai/icons/brands/mistral.svg",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":37},5183,381,{"id":431,"category":4,"slug":432,"name":433,"description":434,"icon":8,"tags":435,"tier":438,"language":24,"aura_score":29,"aura_count":29,"author_id":27,"author_username":28,"energy_cost_per_use":29,"published_at":439,"brand_id":152,"brand_name":153,"brand_icon_light":154,"brand_icon_dark":155,"context_window":137,"capabilities":440,"input_cost":441,"output_cost":442,"image_cost":8,"audio_cost":8,"cache_read_cost":443,"model_type":15,"avg_latency_ms":444,"avg_tps":445,"avg_ttft_ms":446,"avg_image_gen_ms":8,"avg_embed_tps":8,"free_tier_eligible":37,"model_id":432},"f96bd4a8-4c9d-484e-bb22-99af1558c492","kimi-k2.7-code","Kimi K2.7 Code","Moonshot Kimi K2.7 Code is a 1‑trillion‑parameter Mixture‑of‑Experts (MoE) chat model with 32 B activated parameters, featuring a 256 K token context window and a MoonViT vision encoder (400 M parameters). Optimized for coding and agentic tasks, it operates in a forced thinking mode that preserves reasoning across turns. The model achieves 62.0 on the Kimi Code Bench v2, 53.6 on Program Bench, and 76.0 on the MCP Atlas benchmark. It supports text, image, and video inputs and provides function calling capabilities. The model is released under a custom “Other” license.",[210,211,436,363,213,214,14,216,17,18,130,54,22,174,437,16],"k2.7","custom-license","paid","2026-06-16T23:57:16.862507Z",{"vision":36,"function_calling":36,"streaming":36,"structured_output":36,"supports_audio":37,"computer_use":37,"reasoning":36},0.6,3.6,0.16,592,114,620,{"id":448,"category":4,"slug":449,"name":450,"description":451,"icon":8,"tags":452,"tier":23,"language":24,"aura_score":29,"aura_count":29,"author_id":8,"author_username":150,"energy_cost_per_use":29,"published_at":460,"brand_id":461,"brand_name":462,"brand_icon_light":463,"brand_icon_dark":464,"context_window":465,"capabilities":466,"input_cost":40,"output_cost":29,"image_cost":8,"audio_cost":8,"cache_read_cost":8,"model_type":199,"avg_latency_ms":467,"avg_tps":8,"avg_ttft_ms":8,"avg_image_gen_ms":8,"avg_embed_tps":468,"free_tier_eligible":37,"model_id":449},"3d661db7-33de-4d39-bb70-2168f4b21831","bge-m3","BGE M3","BAAI BGE-M3 is a versatile multilingual embedding model supporting dense, sparse, and multi-vector retrieval in a unified architecture. Handles 100+ languages with strong cross-lingual capabilities and flexible retrieval modes for different use cases. Features hybrid retrieval combining dense embeddings for semantic similarity, sparse representations for lexical matching, and multi-vector approaches for fine-grained relevance. Ideal for multilingual search engines, hybrid retrieval systems, and complex information retrieval scenarios requiring multiple matching strategies.",[453,454,455,196,456,172,457,458,459,199],"baai","bge","m3","hybrid-retrieval","sparse","multi-vector","100-languages","2025-12-16T15:32:10.987597Z","f26f99b7-4ef1-425c-bbfe-30bce7a04de0","BAAI","https://cdn.melious.ai/icons/brands/baai_light.svg","https://cdn.melious.ai/icons/brands/baai_dark.svg",8192,{"vision":37,"function_calling":37,"streaming":36,"structured_output":37,"supports_audio":37,"computer_use":37,"reasoning":37},115,1533,35,{"context_window":471,"input_cost_per_million_eur":473,"output_cost_per_million_eur":474,"tps":475,"knowledge_cutoff":477,"available_brands":480},{"min":472,"max":34},128,{"min":29,"max":157},{"min":29,"max":158},{"min":56,"max":476},739,{"min":478,"max":479},"2023-10","2025-04",[481,482,483,484,485,486,487,488,489,490,491,495,496],{"id":461,"name":462,"icon_light":463,"icon_dark":464},{"id":316,"name":317,"icon_light":318,"icon_dark":319},{"id":294,"name":295,"icon_light":8,"icon_dark":8},{"id":31,"name":32,"icon_light":33,"icon_dark":33},{"id":134,"name":135,"icon_light":136,"icon_dark":136},{"id":368,"name":369,"icon_light":370,"icon_dark":370},{"id":424,"name":425,"icon_light":426,"icon_dark":426},{"id":152,"name":153,"icon_light":154,"icon_dark":155},{"id":238,"name":239,"icon_light":240,"icon_dark":241},{"id":178,"name":179,"icon_light":180,"icon_dark":180},{"id":492,"name":493,"icon_light":494,"icon_dark":494},"219f0f8d-9af5-4fed-8b56-adb620dad3aa","Sentence Transformers","https://cdn.melious.ai/icons/brands/sentence_transformers.png",{"id":81,"name":82,"icon_light":83,"icon_dark":84},{"id":331,"name":332,"icon_light":333,"icon_dark":333},{"view":498},"card"],"uses":{"search_params":["page_size","sort","order","q","page","tier","tags","ctx_min","ctx_max","vision","reasoning","fn_call","cutoff_min","cutoff_max","in_min","in_max","out_min","out_max","tps_min","brands","view"],"url":1}}]}