z-ai/glm-5.3-20260816
GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves on GLM-5.2 in coding and in the balance between performance and token efficiency
- Context window
- 1,048,576
- Input modalities
- text
- Output modalities
- text
- Capabilities
- promptCaching, reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 31.7
- OpenRouter Benchmarks: 78.829%
- Vercel AI Gateway: 15.486%
Inference prices
- Cached input tokens (per 1M) : $0.26 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $1.40 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $4.40 / output_tokens_per_million · siliconFlow ↗
z-ai/glm-5.3-flash-20260826
GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context behavior while reducing compute overhead.
- Context window
- 1,048,576
- Input modalities
- text, image, video
- Output modalities
- text
- Capabilities
- multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, videoInput, visionInput
Benchmarks
- Artificial Analysis: 27.8
- OpenRouter Benchmarks: 92.521%
- Vercel AI Gateway: 7.127%
Inference prices
- Cached input tokens (per 1M) : $0.03 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $0.15 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $0.50 / output_tokens_per_million · siliconFlow ↗
deepseek/deepseek-v4-pro-20260813
DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.
- Context window
- 1,048,576
- Input modalities
- text
- Output modalities
- text
- Capabilities
- promptCaching, reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 23.1
- OpenRouter Benchmarks: 91.898%
Inference prices
- Cached input tokens (per 1M) : $0.04 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $1.32 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $3.96 / output_tokens_per_million · siliconFlow ↗
minimax/minimax-m3-20260531
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full a
- Context window
- 1,048,576
- Input modalities
- text, image, video
- Output modalities
- text
- Capabilities
- multimodal, promptCaching, reasoning, streaming, structuredOutputs, toolCalling, videoInput, visionInput
Benchmarks
- Artificial Analysis: 18.9
- OpenCompass LLM: 90.8%
- OpenRouter Benchmarks: 91.659%
Inference prices
- Cached input tokens (per 1M) : $0.06 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $0.30 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $1.20 / output_tokens_per_million · siliconFlow ↗
moonshotai/kimi-k3-20260715
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at navigating large repositories, using tools, debugging, and iterating against ima
- Context window
- 1,048,576
- Input modalities
- text, image, video
- Output modalities
- text
- Capabilities
- audioInput, multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, videoInput, visionInput
Benchmarks
- Artificial Analysis: 17.4
- OpenRouter Benchmarks: 93.07%
- Vercel AI Gateway: 7.458%
Inference prices
- Cached input tokens (per 1M) : $0.27 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $2.70 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $13.50 / output_tokens_per_million · siliconFlow ↗
qwen/qwen3-vl-235b-a22b-thinking
Qwen3-VL-235B-A22B Thinking is a multimodal model that unifies strong text generation with visual understanding across images and video. The Thinking model is optimized for multimodal reasoning in STEM and math....
- Context window
- 131,072
- Input modalities
- text, image
- Output modalities
- text
- Capabilities
- multimodal, structuredOutputs, toolCalling, visionInput
Benchmarks
- Artificial Analysis: 13.4
deepseek/deepseek-v4-pro-20260423
DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding, and long-horizon agent workflows, with strong performance across k
- Context window
- 1,048,576
- Input modalities
- text
- Output modalities
- text
- Capabilities
- audioInput, multimodal, promptCaching, reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 12.6
- Kaggle Benchmarks: 1,387.2
- OpenCompass LLM: 97.1%
- OpenRouter Benchmarks: 91.688%
Inference prices
- Cached input tokens (per 1M) : $0.14 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $1.50 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $3.14 / output_tokens_per_million · siliconFlow ↗
qwen/qwen3-vl-32b-instruct
Qwen3-VL-32B-Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video. With 32 billion parameters, it combines deep visual perception with advanced text...
- Context window
- 131,072
- Input modalities
- text, image
- Output modalities
- text
- Capabilities
- multimodal, structuredOutputs, toolCalling, visionInput
Benchmarks
- Artificial Analysis: 11.9
Inference prices
- Input tokens (per 1M) : $0.20 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $0.60 / output_tokens_per_million · siliconFlow ↗
deepseek/deepseek-r1
DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass....
- Context window
- 64,000
- Input modalities
- text
- Output modalities
- text
- Capabilities
- reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 11.4
- LLM Stats · GitHub archive: 0.959
Inference prices
- Input tokens (per 1M) : $0.50 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $2.18 / output_tokens_per_million · siliconFlow ↗
qwen/qwen3-next-80b-a3b-thinking-2509
Qwen3-Next-80B-A3B-Thinking is a reasoning-first chat model in the Qwen3-Next line that outputs structured “thinking” traces by default. It’s designed for hard multi-step problems; math proofs, code synthesis/debugging, logic, and agentic...
- Context window
- 262,144
- Input modalities
- text
- Output modalities
- text
- Capabilities
- reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 11.2
- HELM Capabilities: 46.683%
- OpenCompass LLM: 89%
qwen/qwen3-30b-a3b-thinking-2507
Qwen3-30B-A3B-Thinking-2507 is a 30B parameter Mixture-of-Experts reasoning model optimized for complex tasks requiring extended multi-step thinking. The model is designed specifically for “thinking mode,” where internal reasoning traces are separated...
- Context window
- 81,920
- Input modalities
- text
- Output modalities
- text
- Capabilities
- structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 9.8
- OpenCompass LLM: 89.7%
qwen/qwen3-235b-a22b-04-28
Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a "thinking" mode for complex reasoning, math, and...
- Context window
- 131,072
- Input modalities
- text
- Output modalities
- text
- Capabilities
- structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 9.5
- LLM Stats · GitHub archive: 0.956
qwen/qwen3-vl-8b-thinking
Qwen3-VL-8B-Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences. It integrates enhanced multimodal alignment and...
- Context window
- 131,072
- Input modalities
- image, text
- Output modalities
- text
- Capabilities
- multimodal, structuredOutputs, toolCalling, visionInput
Benchmarks
qwen/qwen3-8b-04-28
Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between "thinking" mode for math,...
- Context window
- 131,072
- Input modalities
- text
- Output modalities
- text
- Capabilities
- reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 7.3
- OpenCompass LLM: 50.1%
Inference prices
- Input tokens (per 1M) : $0.06 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $0.06 / output_tokens_per_million · siliconFlow ↗
google/gemma-4-31b-it-20260402
Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function calling, and multilingual support across 140+ languages. Strong on
- Context window
- 262,144
- Input modalities
- image, text, video
- Output modalities
- text
- Capabilities
- multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput
Benchmarks
- Artificial Analysis: 4.4
- OpenRouter Benchmarks: 76.44%
Inference prices
- Cached input tokens (per 1M) : $0.25 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $0.75 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $1.00 / output_tokens_per_million · siliconFlow ↗
openai/gpt-oss-120b
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with nativ
- Context window
- 131,072
- Input modalities
- text
- Output modalities
- text
- Capabilities
- promptCaching, reasoning, structuredOutputs, toolCalling
Benchmarks
- Artificial Analysis: 3.8
- HELM Capabilities: 68.833%
- Kaggle Benchmarks: 100%
- LLM Stats · GitHub archive: 0.19
- OpenCompass LLM: 93.4%
- OpenRouter Benchmarks: 76.305%
Inference prices
- Cached input tokens (per 1M) : $0.08 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $0.05 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $0.45 / output_tokens_per_million · siliconFlow ↗
qwen/qwen3.8-2.4t-a95b-20260812
Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion total. It is suited for coding, research, complex reasoning, and agentic workflows.
- Context window
- 1,048,576
- Input modalities
- text
- Output modalities
- text
- Capabilities
- multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput
Benchmarks
- Artificial Analysis: 39.9
- OpenRouter Benchmarks: 91.771%
Inference prices
- Cached input tokens (per 1M) : $0.25 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $2.00 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $6.00 / output_tokens_per_million · siliconFlow ↗
deepseek/deepseek-v4-flash-vision-exp-20260821
DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of [DeepSeek V4 Flash 0731](https://openrouter.ai/deepseek/deepseek-v4-flash-0731) from DeepSeek, adding image understanding while matching the base model on text capabilities including agents, reasoning, and
- Context window
- 1,048,576
- Input modalities
- text, image
- Output modalities
- text
- Capabilities
- multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput
Benchmarks
- Artificial Analysis: 34.8
- OpenRouter Benchmarks: 90.909%
Inference prices
- Cached input tokens (per 1M) : $0.03 / cache_read_tokens_per_million · siliconFlow ↗
- Input tokens (per 1M) : $0.44 / input_tokens_per_million · siliconFlow ↗
- Output tokens (per 1M) : $1.32 / output_tokens_per_million · siliconFlow ↗