{"models":[{"name":"zai-org/GLM-5.3","short_name":"Z.ai: GLM-5.3","short_description":"GLM-5.3 uses the same base model as GLM-5.2, with every gain coming from post-training. It is the most capable open-weights model for coding, with a 50% improvement over GLM-5.2 on Z.ai's in-house Code Bench and open-source SOTA on benchmarks including Terminal Bench 3.0 and Agents' Last Exam.","providers":[{"provider":"tng","tokens_per_second":97,"lifecycle":{"state":"stable"},"mode":"chat","pricing":{"mode":"chat","input":1.5e-6,"output":4.5e-6,"cache_read":3e-7,"price_update":null},"max_total_tokens":230000,"max_input_length":229999}],"default_provider_index":0,"capabilities":["chat","code","english","german","multilingual","reasoning","tools"],"attributes":["flagship","production-stable"],"license":{"name":"GLM-5.3 License","url":"https://huggingface.co/zai-org/GLM-5.3/blob/main/LICENSE"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"zai-org/GLM-5.3","quantization":"nvfp4","default":true,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["low","high","max"],"default_reasoning_effort":"high","supports_vision":false,"supports_response_schema":false},{"name":"zai-org/GLM-5.2","short_name":"Z.ai: GLM-5.2","short_description":"Open-weights model rivaling Claude Opus 4.8 and GPT-5.5.","providers":[{"provider":"tng","tokens_per_second":166,"lifecycle":{"state":"deprecated"},"mode":"chat","pricing":{"mode":"chat","input":1.5e-6,"output":4.5e-6,"cache_read":3e-7,"price_update":null},"max_total_tokens":1048576,"max_input_length":1048575}],"default_provider_index":0,"capabilities":["chat","code","english","german","multilingual","reasoning","tools"],"attributes":["deprecated"],"license":{"name":"MIT","url":"https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"zai-org/GLM-5.2","quantization":"nvfp4","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","high","max"],"default_reasoning_effort":"high","supports_vision":false,"supports_response_schema":false},{"name":"zai-org/GLM-5.3-Flash","short_name":"Z.ai: GLM-5.3-Flash","short_description":"Fast multimodal model, strong for agentic coding","providers":[{"provider":"tng","tokens_per_second":52,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":3e-7,"output":6e-7,"cache_read":6e-8,"price_update":null},"max_total_tokens":1048576,"max_input_length":1048575}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["experimental"],"license":{"name":"MIT","url":"https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"zai-org/GLM-5.3-Flash","quantization":"nvfp4","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["low","high","max"],"default_reasoning_effort":"high","supports_vision":true,"supports_response_schema":false},{"name":"tngtech/DeepSeek-TNG-R1T2-Chimera","short_name":"TNG: DeepSeek R1T2 Chimera","short_description":"Powerful reasoning, for complex coding. Model merge combining weights from DeepSeek-R1, DeepSeek-R1-0528 and DeepSeek-V3-0324.","providers":[{"provider":"tng","tokens_per_second":49,"lifecycle":{"state":"stable"},"mode":"chat","pricing":{"mode":"chat","input":1e-6,"output":3e-6,"cache_read":2e-7,"price_update":null},"max_total_tokens":163840,"max_input_length":163839}],"default_provider_index":0,"capabilities":["chat","code","english","fim","german","multilingual","reasoning","tools"],"attributes":["production-stable"],"license":{"name":"MIT","url":"https://huggingface.co/tngtech/DeepSeek-TNG-R1T2-Chimera/raw/main/LICENSE.DeepSeek"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"tngtech/DeepSeek-TNG-R1T2-Chimera","quantization":"nvfp4","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":null,"default_reasoning_effort":null,"supports_vision":false,"supports_response_schema":false},{"name":"google/diffusiongemma-26B-A4B-it","short_name":"Google: DiffusionGemma","short_description":"Experimental fast multimodal diffusion LLM","providers":[{"provider":"tng","tokens_per_second":3,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":3e-7,"output":8e-7,"cache_read":5e-8,"price_update":null},"max_total_tokens":262144,"max_input_length":262143}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["experimental"],"license":{"name":"Apache-2.0","url":"https://huggingface.co/datasets/choosealicense/licenses/raw/main/markdown/apache-2.0.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["top_p","top_k","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"google/diffusiongemma-26B-A4B-it","quantization":null,"default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","medium"],"default_reasoning_effort":"medium","supports_vision":true,"supports_response_schema":false},{"name":"google/gemma-4-26B-A4B-it","short_name":"Google: Gemma 4 (26B)","short_description":"Lightweight multimodal agent. Well-suited for reasoning, agentic workflows, coding, and multimodal understanding.","providers":[{"provider":"tng","tokens_per_second":67,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":3e-7,"output":8e-7,"cache_read":5e-8,"price_update":null},"max_total_tokens":262144,"max_input_length":262143}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["experimental"],"license":{"name":"Apache-2.0","url":"https://huggingface.co/datasets/choosealicense/licenses/raw/main/markdown/apache-2.0.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"google/gemma-4-26B-A4B-it","quantization":"fp8","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","medium"],"default_reasoning_effort":"medium","supports_vision":true,"supports_response_schema":false},{"name":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B","short_name":"NVIDIA: Nemotron 3.5 Super VL","short_description":"Experimental, reasoning, and vision-capable","providers":[{"provider":"tng","tokens_per_second":22,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":1.5e-7,"output":5e-7,"cache_read":2e-8,"price_update":null},"max_total_tokens":262144,"max_input_length":262143}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["experimental"],"license":{"name":"OpenMDW-1.1","url":"https://openmdw.ai/license/1-1/"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B","quantization":null,"default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","medium"],"default_reasoning_effort":"medium","supports_vision":true,"supports_response_schema":false},{"name":"openai/gpt-oss-120b","short_name":"OpenAI: GPT OSS","short_description":"General purpose, high reasoning. Open-weight model for powerful reasoning, agentic tasks, and versatile developer use.","providers":[{"provider":"tng","tokens_per_second":190,"lifecycle":{"state":"stable"},"mode":"chat","pricing":{"mode":"chat","input":1.5e-7,"output":6e-7,"cache_read":5e-8,"price_update":null},"max_total_tokens":131072,"max_input_length":131071}],"default_provider_index":0,"capabilities":["chat","code","english","german","multilingual","reasoning","tools"],"attributes":["production-stable","long-term-stable"],"license":{"name":"Apache-2.0","url":"https://huggingface.co/datasets/choosealicense/licenses/raw/main/markdown/apache-2.0.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"openai/gpt-oss-120b","quantization":"mxfp4","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":false,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["low","medium","high"],"default_reasoning_effort":"medium","supports_vision":false,"supports_response_schema":false},{"name":"Qwen/Qwen3.6-35B-A3B-FP8","short_name":"Qwen: Qwen3.6 (small)","short_description":"Lightweight multimodal agent. Unified vision-language foundation with global linguistic coverage.","providers":[{"provider":"tng","tokens_per_second":189,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":2.5e-7,"output":1.5e-6,"cache_read":5e-8,"price_update":null},"max_total_tokens":262144,"max_input_length":262143}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":[],"license":{"name":"Apache-2.0","url":"https://huggingface.co/datasets/choosealicense/licenses/raw/main/markdown/apache-2.0.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"Qwen/Qwen3.6-35B-A3B-FP8","quantization":"fp8","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","medium"],"default_reasoning_effort":"medium","supports_vision":true,"supports_response_schema":false},{"name":"deepseek-ai/DeepSeek-V4.1-Flash-1017","short_name":"DeepSeek: V4.1-Flash (1017)","short_description":"Latest checkpoint (1017) of the experimental multimodal flash variant of DeepSeek-V4.1: a 552B-parameter MoE with aggressive KV cache compression, supporting contexts of up to one million tokens.","providers":[{"provider":"tng","tokens_per_second":22,"lifecycle":{"state":"experimental"},"mode":"chat","pricing":{"mode":"chat","input":3e-7,"output":6e-7,"cache_read":6e-8,"price_update":null},"max_total_tokens":1048576,"max_input_length":1048575}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["experimental"],"license":{"name":"MIT","url":"https://choosealicense.com/licenses/mit/"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","quantization":"fp8","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["low","high","max"],"default_reasoning_effort":"high","supports_vision":true,"supports_response_schema":false},{"name":"Qwen/Qwen3.8-27B","short_name":"Qwen: Qwen3.8 (small)","short_description":"Compact vision-language agent. 27B dense model with image and video understanding, flexible reasoning control, 1M context.","providers":[{"provider":"tng","tokens_per_second":75,"lifecycle":{"state":"stable"},"mode":"chat","pricing":{"mode":"chat","input":4e-7,"output":2.7e-6,"cache_read":8e-8,"price_update":null},"max_total_tokens":1000000,"max_input_length":999999}],"default_provider_index":0,"capabilities":["chat","code","english","german","images","multilingual","reasoning","tools"],"attributes":["production-stable"],"license":{"name":"Apache-2.0","url":"https://huggingface.co/datasets/choosealicense/licenses/raw/main/markdown/apache-2.0.md"},"supported_endpoints":["/v1/chat/completions","/anthropic/v1/messages","/v1/responses"],"model_parameters":["temperature","top_p","top_k","seed","max_tokens","max_completion_tokens","logprobs","top_logprobs","response_format","stop","stream","reasoning_effort","tools","tool_choice"],"hugging_face_id":"Qwen/Qwen3.8-27B","quantization":"nvfp4","default":false,"mode":"chat","max_output_tokens":null,"supports_function_calling":true,"supports_parallel_function_calling":true,"supports_tool_choice":true,"supports_reasoning":true,"supported_reasoning_effort":["none","low","medium","xhigh"],"default_reasoning_effort":"medium","supports_vision":true,"supports_response_schema":false}],"default_provider":"tng"}