Explore 249+ AI models from 49 providers. Filter by capability, tier, and pricing to find the right model.
Showing all 249 models
Moonshot AI
Moonshot AI's frontier multimodal MoE model with 1T total parameters (32B active). Tops SWE-bench and AIME 2025 benchmarks.
Input
$0.45/M
Output
$2.20/M
Context
256K
Meta
Meta's open-source model matching GPT-4 class performance at 70B parameters.
Input
$0.18/M
Output
$0.18/M
Context
128K
OpenAI
OpenAI's second-generation video synthesis model capable of producing cinematic-quality videos up to 60 seconds long with synchronized audio. Built on an advanced Diffusion Transformer (DiT) architecture, Sora 2 excels at complex scene composition, realistic physics simulation, and coherent multi-character narratives with natural dialogue and ambient sound.
Input
$5.00/M
Output
$100.00/M
OpenAI
OpenAI's flagship model replacing GPT-4o and o3. Achieves 94.6% on AIME 2025 and 74.9% on SWE-bench. Multimodal with thinking capabilities.
Input
$5.00/M
Output
$15.00/M
Context
197K
OpenAI
Expanded GPT-5 with 400K context and 128K max output. Near-perfect 100% on AIME 2025 math benchmark.
Input
$8.00/M
Output
$24.00/M
Context
400K
Google DeepMind's flagship video generation model that natively produces joint audio-visual output in a single pass. Veo 3 leverages a Latent Diffusion Transformer to generate high-fidelity clips with synchronized dialogue, sound effects, and ambient audio without requiring a separate audio model. It demonstrates strong physical understanding and prompt adherence across diverse cinematic styles.
Input
$5.00/M
Output
$150.00/M
OpenAI
OpenAI's agentic coding model with context compaction and long-horizon task completion. First model in the GPT-5 Codex series.
Input
$1.75/M
Output
$14.00/M
Context
400K
OpenAI
OpenAI's most capable coding model combining Codex and GPT-5 training stacks. Agentic coding, research, and tool use with 77.3% on Terminal-Bench 2.0.
Input
$2.00/M
Output
$16.00/M
Context
400K
Anthropic
Anthropic's strongest reasoning and coding model. 80.8% on SWE-bench Verified, 1M context (beta), agent teams, and extended thinking.
Input
$5.00/M
Output
$25.00/M
Context
1.0M
Anthropic
Matches Opus 4.6 on most benchmarks at 1/5 the cost. 79.6% on SWE-bench, 1M context, computer use, and design capabilities.
Input
$3.00/M
Output
$15.00/M
Context
1.0M
Google's frontier-class model at Flash-level latency and cost. 90.4% on GPQA Diamond, 78% on SWE-bench, 1M context window.
Input
$0.50/M
Output
$3.00/M
Context
1.0M
Google's most capable model. 94.3% on GPQA Diamond, 80.6% on SWE-bench, 77.1% on ARC-AGI-2. #1 on 12 of 18 tracked benchmarks.
Input
$2.00/M
Output
$12.00/M
Context
1.0M
DeepSeek
DeepSeek's 1T parameter coding-focused model with 1M+ context. Three architectural innovations: Manifold-Constrained Hyper-Connections, Engram memory, Sparse Attention.
Input
$0.10/M
Output
$0.40/M
Context
1.0M
Alibaba/Qwen
Alibaba's open-weight hybrid MoE model with 512 experts and 17B active parameters. Natively multimodal with 201 language support. Top scores on GPQA and SWE-bench.
Input
$0.15/M
Output
$1.00/M
Context
256K
Meta
Meta's latest open-source MoE model with 17B active parameters and industry-leading 10M token context.
Input
$0.15/M
Output
$0.60/M
Context
10.5M
Anthropic
Flagship Opus release with major improvements in coding and workplace productivity tasks. Predecessor to Opus 4.6.
Input
$15.00/M
Output
$75.00/M
Context
1.0M
Runway
Runway's latest flagship model built on a novel Autoregressive-to-Diffusion (A2D) hybrid architecture that first plans scene structure autoregressively and then renders frames through diffusion. Gen-4.5 achieves state-of-the-art temporal coherence, photorealistic detail, and nuanced control over motion dynamics, lighting, and artistic style.
Input
$5.00/M
Output
$130.00/M
xAI
xAI's frontier model trained on Colossus supercluster. Real-time data access and strong reasoning.
Input
$3.00/M
Output
$15.00/M
Context
131K
Amazon
Most intelligent Amazon model for complex multi-step reasoning and agentic workflows.
Input
$4.00/M
Output
$12.00/M
Context
1.0M
Most powerful Gemini model with native multimodal understanding. Supports adjustable reasoning depth via thinking_level parameter.
Input
$3.50/M
Output
$10.50/M
Context
1.0M
Kuaishou
The latest entry in Kuaishou's Kling series, introducing multi-shot sequence generation and an AI Director mode that automatically plans camera angles, transitions, and pacing. Kling 3.0 produces broadcast-quality video with native audio synthesis, making it suitable for short-form content creation, advertising, and virtual production workflows.
Input
$3.00/M
Output
$60.00/M
Meta
Meta's powerful open-source MoE model with 400B total params and 1M context window.
Input
$0.50/M
Output
$2.00/M
Context
1.0M
Anthropic
Anthropic's most powerful model. Top-tier performance on coding, analysis, and complex reasoning tasks.
Input
$15.00/M
Output
$75.00/M
Context
200K
Anthropic
Anthropic's best balance of intelligence and speed. Excellent for production workloads.
Input
$3.00/M
Output
$15.00/M
Context
200K
ByteDance
ByteDance's unified multimodal generation model that handles video, audio, and image synthesis within a single architecture. Seedance 2.0 produces highly coherent audiovisual content with strong temporal consistency, supporting diverse creative workflows from music video generation to product advertisement creation with synchronized narration and effects.
Input
$3.00/M
Output
$70.00/M
OpenAI
OpenAI's reasoning model with chain-of-thought capabilities for complex problem solving.
Input
$15.00/M
Output
$60.00/M
Context
200K
Google's fastest multimodal model with native tool use and advanced agentic capabilities.
Input
$0.10/M
Output
$0.40/M
Context
1.0M
OpenAI
OpenAI's most powerful reasoning model with breakthrough performance on math and coding benchmarks.
Input
$10.00/M
Output
$40.00/M
Context
200K
Alibaba/Qwen
Specialized code model trained on 7.5T tokens (70% code). Supports 100+ programming languages and agentic workflows.
Input
$0.30/M
Output
$0.60/M
Context
262K
Alibaba/Qwen
Most capable open VLM rivaling GPT-5 across multimodal benchmarks. Strong reasoning and agentic capabilities.
Input
$0.30/M
Output
$0.60/M
Context
128K
MiniMax
Achieves 80.2% on SWE-Bench Verified matching Opus 4.6 at 1/20th cost. First on Multi-SWE-Bench at 51.3%.
Input
$0.25/M
Output
$0.75/M
Context
128K
DeepSeek
DeepSeek's open-source MoE model rivaling frontier models at a fraction of the cost.
Input
$0.27/M
Output
$1.10/M
Context
128K
DeepSeek
DeepSeek's reasoning model with transparent chain-of-thought. Open-source and highly competitive.
Input
$0.55/M
Output
$2.19/M
Context
128K
OpenAI
OpenAI's native image generation capability integrated directly into GPT-4o, enabling conversational image creation and iterative editing through natural language. GPT Image 1 excels at accurate text rendering within images, complex multi-element compositions, and faithful adherence to detailed prompts across photorealistic, illustrative, and artistic styles.
Input
$10.00/M
Output
$40.00/M
DeepSeek
Hybrid model combining V3 and R1 strengths. Improved reasoning with RL techniques from R1.
Input
$0.27/M
Output
$1.10/M
Context
128K
Google's fast and cost-efficient thinking model with strong reasoning capabilities.
Input
$0.15/M
Output
$0.60/M
Context
1.0M
Google's most capable thinking model with breakthrough performance on reasoning and coding.
Input
$1.25/M
Output
$10.00/M
Context
1.0M
Google DeepMind's fourth-generation image synthesis model capable of producing images up to 2K resolution with exceptional photorealism and compositional accuracy. Imagen 4 includes SynthID watermarking by default for responsible AI deployment, supports advanced inpainting and outpainting, and demonstrates industry-leading performance on text rendering and spatial reasoning tasks.
Input
$4.00/M
Output
$20.00/M
OpenAI
OpenAI's latest GPT-4 series model with improved coding, instruction following, and long context.
Input
$2.00/M
Output
$8.00/M
Context
1.0M
Black Forest Labs
Black Forest Labs' flagship commercial image generation model with 32 billion parameters, delivering up to 4-megapixel resolution output with exceptional detail and prompt fidelity. FLUX.2 Pro achieves state-of-the-art results in photorealism, typography rendering, and complex scene composition, making it a top choice for professional creative applications.
Input
$3.00/M
Output
$30.00/M
OpenAI
OpenAI's efficient reasoning model, optimized for speed while maintaining strong analytical capabilities.
Input
$1.10/M
Output
$4.40/M
Context
200K
Midjourney
Midjourney's seventh major model release featuring 12 billion parameters and expanded multimodal capabilities including short video clip generation alongside its renowned image synthesis. V7 delivers dramatically improved coherence, photorealism, and artistic range, with enhanced understanding of spatial relationships, lighting, and material properties across diverse visual styles.
Input
$5.00/M
Output
$50.00/M
OpenAI
OpenAI's most advanced multimodal model. Excels at text, vision, and audio tasks with fast response times.
Input
$2.50/M
Output
$10.00/M
Context
128K
Alibaba/Qwen
Alibaba's flagship open-source model. Competitive with Llama 3.1 405B at a fraction of the size.
Input
$0.30/M
Output
$0.30/M
Context
128K
Alibaba/Qwen
Strong mid-range open-source model from Alibaba with broad capabilities.
Input
$0.08/M
Output
$0.08/M
Context
128K
Alibaba/Qwen
Efficient open-source model balancing capability and speed.
Input
$0.05/M
Output
$0.05/M
Context
128K
Alibaba/Qwen
Compact open-source model for edge deployment and fine-tuning.
Input
$0.03/M
Output
$0.03/M
Context
32K
Alibaba/Qwen
Alibaba's open-source coding specialist. Matches GPT-4o on code benchmarks.
Input
$0.08/M
Output
$0.08/M
Context
128K
Alibaba/Qwen
Compact open-source coding model with impressive code generation capabilities.
Input
$0.03/M
Output
$0.03/M
Context
128K
Alibaba/Qwen
Alibaba's open-source vision-language model with video understanding capabilities.
Input
$0.40/M
Output
$0.40/M
Context
32K
Alibaba/Qwen
Alibaba's open-source reasoning model with transparent chain-of-thought. Competitive with o1-mini.
Input
$0.10/M
Output
$0.30/M
Context
32K
Alibaba/Qwen
Alibaba's large-scale open-source MoE model with thinking mode support.
Input
$0.20/M
Output
$0.60/M
Context
128K
Microsoft
Microsoft's 14B open-source model with training innovations that punch above its weight class.
Input
$0.04/M
Output
$0.04/M
Context
16K
Microsoft
Microsoft's compact open-source model with 128K context. Great for on-device inference.
Input
$0.01/M
Output
$0.01/M
Context
128K
Microsoft
Microsoft's open-source MoE model with 42B total params and only 6.6B active.
Input
$0.06/M
Output
$0.06/M
Context
128K
Microsoft
Microsoft's 14B open-source model with 128K context and strong reasoning capabilities.
Input
$0.04/M
Output
$0.04/M
Context
128K
Microsoft
Microsoft's instruction-tuned MoE model based on Mixtral. Strong on complex reasoning tasks.
Input
$0.65/M
Output
$0.65/M
Context
66K
NVIDIA
NVIDIA's optimized Llama 3.1 variant with custom reward model training.
Input
$0.18/M
Output
$0.18/M
Context
128K
NVIDIA
NVIDIA's large open-source model trained for synthetic data generation.
Input
$1.20/M
Output
$1.20/M
Context
4K
AI21 Labs
AI21's hybrid SSM-Transformer model with 256K context. Novel Mamba architecture.
Input
$2.00/M
Output
$8.00/M
Context
256K
AI21 Labs
Compact version of Jamba with hybrid SSM-Transformer architecture.
Input
$0.20/M
Output
$0.40/M
Context
256K
TII
TII's largest open-source model. One of the first truly open 180B parameter models.
Input
$0.80/M
Output
$0.80/M
Context
2K
TII
TII's efficient open-source model with multimodal capabilities.
Input
$0.04/M
Output
$0.04/M
Context
8K
01.AI
01.AI's fast inference model with strong performance across benchmarks.
Input
$0.30/M
Output
$0.30/M
Context
16K
01.AI
01.AI's open-source 34B model with strong bilingual (English/Chinese) capabilities.
Input
$0.10/M
Output
$0.10/M
Context
4K
01.AI
01.AI's frontier closed-source model with top-tier multilingual performance.
Input
$3.00/M
Output
$9.00/M
Context
32K
Shanghai AI Lab
Open-source model with 1M context from Shanghai AI Lab. Strong coding and math skills.
Input
$0.06/M
Output
$0.06/M
Context
1.0M
Shanghai AI Lab
Open-source vision-language model with strong image understanding capabilities.
Input
$0.08/M
Output
$0.08/M
Context
8K
Stability AI
Stability AI's open-source language model with multilingual support.
Input
$0.04/M
Output
$0.04/M
Context
4K
Allen AI
Fully open-source model from Allen AI with open training data, code, and weights.
Input
$0.04/M
Output
$0.04/M
Context
4K
BigCode
Open-source code model from BigCode/HuggingFace trained on The Stack v2.
Input
$0.04/M
Output
$0.04/M
Context
16K
Snowflake
Snowflake's open-source enterprise MoE model optimized for SQL and business tasks.
Input
$0.30/M
Output
$0.30/M
Context
4K
Databricks
Databricks' open-source MoE model with strong code and reasoning capabilities.
Input
$0.75/M
Output
$0.75/M
Context
32K
Zhipu AI
Zhipu AI's flagship model with strong Chinese and English bilingual capabilities.
Input
$1.00/M
Output
$3.00/M
Context
128K
Zhipu AI
Zhipu AI's latest open-weight MoE model with interleaved thinking and state-of-the-art coding performance.
Input
$0.50/M
Output
$1.50/M
Context
200K
OpenAI
OpenAI's cost-efficient reasoning model with multimodal input, strong math and coding performance at a fraction of o3 pricing.
Input
$1.10/M
Output
$4.40/M
Context
200K
OpenAI
OpenAI's highest-quality reasoning model with extended compute for complex scientific and mathematical problems.
Input
$20.00/M
Output
$80.00/M
Context
200K
Google's ultra-efficient model offering better performance than Gemini 1.5 Flash at the same cost point.
Input
$0.07/M
Output
$0.30/M
Context
1.0M
Mistral AI
Mistral's open-weight 675B MoE model with 41B active parameters, multimodal input, and 256K context.
Input
$0.50/M
Output
$1.50/M
Context
256K
Mistral AI
Mistral's mid-tier model offering 90% of Claude Sonnet quality at significantly lower cost.
Input
$0.40/M
Output
$2.00/M
Context
131K
Mistral AI
Compact 24B model with image understanding, 128K context, and Apache 2.0 license.
Input
$0.10/M
Output
$0.30/M
Context
128K
Mistral AI
Mistral's specialized code model supporting 80+ languages with 256K context and fill-in-the-middle capability.
Input
$0.30/M
Output
$0.90/M
Context
256K
xAI
xAI's efficient reasoning model with fast inference and competitive performance at budget pricing.
Input
$0.30/M
Output
$0.50/M
Context
131K
Microsoft
Microsoft's 3.8B parameter model with 128K context, strong reasoning capability for on-device deployment.
Input
$0.01/M
Output
$0.01/M
Context
128K
Microsoft
Microsoft's 5.6B compact model unifying text, vision, and speech in a single architecture.
Input
$0.02/M
Output
$0.02/M
Context
128K
Microsoft
Chain-of-thought reasoning variant of Phi-4, competitive with much larger models on math and logic tasks.
Input
$0.04/M
Output
$0.04/M
Context
32K
DeepSeek
Unified reasoning and non-reasoning model that merges DeepSeek-V3 and R1 capabilities into a single architecture.
Input
$0.28/M
Output
$0.42/M
Context
128K
DeepSeek
R1's reasoning capability distilled into a Llama 3.1 70B architecture for efficient deployment.
Input
$0.18/M
Output
$0.18/M
Context
128K
Cohere
Cohere's 111B parameter model supporting 23 languages with enterprise tool use and 256K context.
Input
$2.50/M
Output
$10.00/M
Context
256K
Alibaba/Qwen
Alibaba's dense 32B model with dual thinking/non-thinking modes and strong reasoning performance.
Input
$0.08/M
Output
$0.20/M
Context
131K
Alibaba/Qwen
Compact 8B model from the Qwen3 family with thinking mode support and strong efficiency for on-device use.
Input
$0.03/M
Output
$0.06/M
Context
131K
Alibaba/Qwen
Ultra-efficient MoE model with 128 experts and only 3.3B active parameters, ideal for cost-sensitive deployments.
Input
$0.02/M
Output
$0.04/M
Context
131K
xAI
xAI's 4-agent parallel collaboration system with rapid learning architecture and medical document analysis. Beta release.
Input
$3.00/M
Output
$15.00/M
Context
131K
Cohere
Cohere's compact multilingual model supporting 70+ languages. Runs on consumer devices including phones. Outperforms Gemma3-4B in 46/61 languages.
Input
$0.01/M
Output
$0.01/M
Context
32K
TII
Compact Falcon model for resource-constrained deployments with strong reasoning.
Input
$0.04/M
Output
$0.08/M
Context
32K
TII
Smallest Falcon model for edge inference and mobile deployment.
Input
$0.02/M
Output
$0.04/M
Context
32K
Allen AI
Fully open model with all components public: data, code, weights, and checkpoints. Instruct, Think, and RL Zero variants.
Input
$0.25/M
Output
$0.75/M
Context
128K
Allen AI
Outperforms Llama 3.1 8B. Everything released: training data, weights, code, recipes, and checkpoints.
Input
$0.07/M
Output
$0.14/M
Context
128K
Allen AI
Open multimodal model for visual understanding, image captioning, and visual question answering.
Input
$0.40/M
Output
$1.20/M
Context
128K
Zhipu AI
Zhipu's largest text generation model at 754B parameters.
Input
$2.00/M
Output
$6.00/M
Context
256K
Zhipu AI
Vision-language MoE model with superior performance at lower inference cost.
Input
$0.15/M
Output
$0.30/M
Context
128K
Zhipu AI
Open-source video generation model creating 6-second clips at 720x480. Supports LoRA fine-tuning.
Input
$0.05/M
Output
$0.05/M
BAAI
State-of-the-art multimodal embedding model for visual search applications.
Input
$0.02/M
Output
$0.02/M
Context
8K
An enhanced iteration of Google DeepMind's Veo series that produces 8-second clips that can be seamlessly extended up to 148 seconds through iterative generation. Veo 3.1 improves temporal consistency over long sequences, delivers higher resolution output, and refines audio synchronization for extended storytelling and commercial content production.
Input
$3.00/M
Output
$80.00/M
Runway
Runway's high-performance video generation model optimized for professional content creation at up to 4K resolution. Gen-4 Turbo maintains consistent characters and environments across shots, supports detailed camera control, and delivers studio-grade output with dramatically reduced generation times compared to its predecessors.
Input
$5.00/M
Output
$120.00/M
Kuaishou
Kuaishou's advanced video generation model capable of simultaneous audio-visual synthesis, producing clips with fully synchronized dialogue, music, and environmental sounds. Kling 2.6 excels at generating realistic human motion, facial expressions, and complex multi-object interactions while maintaining strong temporal consistency across extended sequences.
Input
$2.00/M
Output
$40.00/M
Pika Labs
Pika Labs' video generation model focused on ultra-realistic output with enhanced physics simulation. Pika 2.5 handles complex material interactions such as fluid dynamics, cloth draping, and particle effects with high fidelity. Its intuitive prompt interface and style controls make it accessible for creators seeking photorealistic short-form video content.
Input
$2.00/M
Output
$50.00/M
Luma AI
Luma AI's video generation model delivering native 1080p output at 4x faster inference speeds than previous versions, with optional 4K upscaling. Ray3.14 specializes in photorealistic 3D-aware video synthesis with strong spatial understanding, making it particularly effective for product visualization, architectural walkthroughs, and immersive content creation.
Input
$2.00/M
Output
$45.00/M
MiniMax
MiniMax's Hailuo 2.3 video model combines photorealistic rendering with versatile style support including anime, watercolor, and cinematic looks. It features advanced motion control, accurate lip-sync for dialogue scenes, and sophisticated lighting effects that adapt dynamically to scene content and camera movement.
Input
$2.00/M
Output
$50.00/M
Lightricks
Lightricks' open-source video generation model capable of producing native 4K video at 50 frames per second with clips up to 20 seconds in length. LTX-2 includes native audio synthesis and offers full model weights under a permissive license, making it a leading choice for researchers and developers building custom video generation pipelines.
Input
Free/M
Output
Free/M
Alibaba/Qwen
Alibaba's open-source video generation model that achieved the number one ranking on the VBench video quality benchmark upon release. With 14 billion parameters, Wan 2.1 demonstrates exceptional prompt adherence, temporal consistency, and visual quality across diverse content types, establishing a new baseline for open-weight video synthesis models.
Input
Free/M
Output
Free/M
Alibaba/Qwen
The successor to Wan 2.1, this open-source model introduces a Mixture-of-Experts flow-matching architecture with approximately 27 billion total parameters and 14 billion active during inference. Wan 2.2 delivers significantly improved motion quality, fine-grained detail, and extended generation lengths while maintaining the accessibility of fully open weights.
Input
Free/M
Output
Free/M
Tencent
Tencent's open-source video generation model with 8.3 billion parameters, featuring a novel Spatial-Temporal Self-Attention (SSTA) mechanism for improved temporal coherence. HunyuanVideo 1.5 supports diverse aspect ratios, variable frame rates, and extended clip durations, making it a versatile foundation model for the open-source video generation community.
Input
Free/M
Output
Free/M
Stability AI
Stability AI's specialized model for 4D novel-view video synthesis, generating temporally consistent multi-angle video from a single input clip or image. Stable Video 4D 2.0 enables creators to produce orbiting camera paths, bullet-time effects, and 3D-aware video transformations that maintain geometric and photometric coherence throughout the sequence.
Input
$2.00/M
Output
$40.00/M
OpenAI
An optimized successor to GPT Image 1 that delivers 20% lower cost and 4x faster generation while maintaining equivalent visual quality. GPT Image 1.5 introduces improved batch processing, enhanced style consistency for multi-image projects, and refined detail handling for professional design and marketing workflows.
Input
$8.00/M
Output
$32.00/M
OpenAI
A cost-efficient variant of OpenAI's image generation model offering 54-70% lower pricing while retaining strong prompt adherence and visual quality for standard use cases. GPT Image 1 Mini is optimized for high-volume applications such as e-commerce product imagery, social media content, and rapid prototyping where speed and cost matter more than maximum fidelity.
Input
$2.50/M
Output
$8.00/M
A multimodal extension of Google's Gemini 2.5 Flash model that adds native image generation and editing capabilities alongside text understanding. This model enables conversational image creation, iterative visual refinement, and combined text-image output within a single unified interface, making it particularly effective for design iteration and creative brainstorming workflows.
Input
$0.15/M
Output
$30.00/M
Black Forest Labs
The open-weights development version of FLUX.2 with the same 32 billion parameter architecture as the Pro variant, released for non-commercial research and experimentation. FLUX.2 Dev provides researchers full access to model weights for fine-tuning, distillation, and architectural exploration while delivering near-Pro-level quality for academic and personal projects.
Input
Free/M
Output
Free/M
Adobe
Adobe's latest commercially safe image generation model trained exclusively on licensed and public domain content, delivering photorealistic output at native 4-megapixel resolution. Firefly Image 5 integrates deeply with Adobe Creative Cloud, offering advanced composition controls, style references, and seamless editing workflows within Photoshop and Illustrator.
Input
$4.00/M
Output
$35.00/M
Adobe
Adobe's fourth-generation Firefly image model offering improved quality, faster generation, and enhanced creative controls compared to its predecessors. Firefly Image 4 provides robust structure references, style transfer, and generative fill capabilities, all trained on Adobe's commercially licensed dataset to ensure IP safety for enterprise and professional use.
Input
$3.00/M
Output
$25.00/M
Stability AI
Stability AI's largest open-source image generation model built on the Multimodal Diffusion Transformer (MMDiT) architecture. SD 3.5 Large delivers high-quality results across photorealistic and artistic styles with strong prompt adherence, accurate text rendering, and diverse composition capabilities, available under an open license for both research and commercial use.
Input
$0.50/M
Output
$6.50/M
Ideogram
Ideogram's third-generation model combining exceptional photorealism with industry-leading text rendering accuracy within generated images. Ideogram 3.0 handles complex typography, logos, signs, and handwritten text with remarkable fidelity, making it the preferred choice for design professionals working on brand assets, marketing materials, and content requiring reliable in-image text.
Input
$2.00/M
Output
$20.00/M
NVIDIA
Input
Free/M
Output
Free/M
Recraft
Recraft's flagship image generation model that achieved the number one ranking on the HuggingFace text-to-image leaderboard, with native support for both raster and vector output formats. Recraft V3 excels at brand-consistent design, offering precise color palette control, style locking, and batch generation capabilities that make it uniquely suited for professional design systems.
Input
$2.00/M
Output
$20.00/M
Microsoft
Microsoft's first in-house image generation model developed by Microsoft AI, designed for integration across Microsoft's product ecosystem including Designer, Copilot, and Bing Image Creator. MAI-Image-1 focuses on safety, controllability, and consistent quality, with built-in content filtering and provenance metadata for responsible enterprise deployment.
Input
$2.00/M
Output
$15.00/M
MiniMax
World's first open-weight large-scale hybrid-attention reasoning model. Natively supports 1M token context.
Input
$0.30/M
Output
$0.90/M
Context
1.0M
Tencent
One of the largest open-source MoE models. Supports text sequences up to 256K tokens.
Input
$0.50/M
Output
$1.50/M
Context
256K
Tencent
World's largest open-source text-to-image model using MoE architecture with 64 experts.
Input
$0.03/M
Output
$0.03/M
Shanghai AI Lab
State-of-the-art open multimodal LLM scoring 72.2 on MMMU. New record among open MLLMs.
Input
$0.40/M
Output
$1.20/M
Context
128K
Shanghai AI Lab
Latest InternLM series model. Efficient for research and application development.
Input
$0.07/M
Output
$0.14/M
Context
128K
Shanghai AI Lab
Advanced vision-language model with improved document and chart understanding capabilities.
Input
$0.40/M
Output
$1.20/M
Context
128K
01.AI
Mid-size Yi model with enhanced inference speed for extended prompts.
Input
$0.10/M
Output
$0.20/M
Context
128K
01.AI
Compact Yi model offering strong reasoning at minimal resource requirements.
Input
$0.06/M
Output
$0.12/M
Context
128K
01.AI
Vision-language Yi model for image understanding and visual question answering.
Input
$0.30/M
Output
$0.60/M
Context
16K
BigCode
Compact code model trained on 4T+ tokens and 600+ languages from The Stack v2.
Input
$0.03/M
Output
$0.06/M
Context
16K
BigCode
Mid-size code model matching CodeLlama 13B quality at half the parameters.
Input
$0.07/M
Output
$0.14/M
Context
16K
Stability AI
Lightweight language model for on-device inference and resource-constrained environments.
Input
$0.02/M
Output
$0.04/M
Context
4K
Stability AI
Mid-size Stable Diffusion optimized for consumer GPUs and edge devices.
Input
$0.02/M
Output
$0.02/M
Black Forest Labs
Fastest FLUX model generating and editing images in under one second. Fully open under Apache 2.0.
Input
$0.01/M
Output
$0.01/M
Black Forest Labs
Fast open-source text-to-image model with 4-step generation. Apache 2.0 licensed.
Input
$0.02/M
Output
$0.02/M
Black Forest Labs
Premium text-to-image model with highest technical quality and 4.5-second generation.
Input
$0.05/M
Output
$0.05/M
Meta
Safety classification model for detecting unsafe content in LLM inputs and outputs.
Input
$0.05/M
Output
$0.05/M
Context
128K
Smallest Gemma 2 model for efficient text processing on consumer hardware.
Input
$0.02/M
Output
$0.04/M
Context
8K
Baichuan
Premier Chinese LLM specializing in law, finance, medicine, and classical literature.
Input
$1.50/M
Output
$4.50/M
Context
128K
LG AI Research
Korean sovereign AI model using MoE with hybrid attention for reduced computation.
Input
$0.25/M
Output
$0.75/M
Context
128K
LG AI Research
Ultra-compact Korean AI model for on-device and mobile deployment.
Input
$0.02/M
Output
$0.04/M
Context
128K
Upstage
Agentic reasoning-focused model matching larger rivals. Strong multilingual capabilities.
Input
$0.20/M
Output
$0.60/M
Context
128K
BAAI
Most popular open embedding model. Multi-functionality, multi-linguality, multi-granularity in one model.
Input
$0.01/M
Output
$0.01/M
Context
8K
OpenAI
Gold standard speech recognition model supporting 99+ languages. 1.55B parameter encoder-decoder architecture.
Input
$0.0060/M
Output
$0.0060/M
OpenAI
Speed-optimized Whisper variant with 6x faster inference at 809M parameters.
Input
$0.0030/M
Output
$0.0030/M
OpenAI
OpenAI's image generation model excelling at precision, complex prompts, and readable text rendering within images.
Input
$0.04/M
Output
$0.04/M
Anthropic
High-intelligence Sonnet model with 1M token context window. Strong balance of performance and cost.
Input
$3.00/M
Output
$15.00/M
Context
1.0M
Anthropic
Fastest and most cost-efficient Claude model designed for high-throughput, low-latency applications.
Input
$0.80/M
Output
$4.00/M
Context
200K
Specialized reasoning model designed for science, research, and complex engineering challenges.
Input
$5.00/M
Output
$15.00/M
Context
1.0M
Smallest Gemma 3 model for edge and mobile deployment. Text-only with 128K context.
Input
$0.02/M
Output
$0.02/M
Context
128K
Open vision-language model for image captioning, visual QA, and OCR tasks. Built on Gemma 2 backbone.
Input
$0.30/M
Output
$0.60/M
Context
8K
Mid-size PaliGemma for efficient vision-language tasks. Strong OCR and document understanding.
Input
$0.15/M
Output
$0.30/M
Context
8K
Alibaba/Qwen
Dense model with hybrid thinking/non-thinking modes. Seamless switching between complex reasoning and general dialogue.
Input
$0.20/M
Output
$0.60/M
Context
128K
Alibaba/Qwen
Compact Qwen3 model with hybrid reasoning for edge deployment and resource-constrained environments.
Input
$0.05/M
Output
$0.15/M
Context
128K
Alibaba/Qwen
Lightweight Qwen3 model for on-device AI applications with reasoning capability.
Input
$0.02/M
Output
$0.06/M
Context
128K
Alibaba/Qwen
Smallest Qwen3 model designed for ultra-lightweight deployment and edge inference.
Input
$0.01/M
Output
$0.03/M
Context
32K
Alibaba/Qwen
Compact vision-language model excelling at video and image analysis. Top small multimodal model on Hugging Face.
Input
$0.10/M
Output
$0.30/M
Context
128K
Alibaba/Qwen
Audio-language model for speech recognition, audio understanding, and music analysis.
Input
$0.10/M
Output
$0.30/M
Context
128K
Alibaba/Qwen
Smallest Qwen VL model for lightweight vision-language tasks on constrained hardware.
Input
$0.04/M
Output
$0.12/M
Context
128K
Alibaba/Qwen
Math-specialized model with step-by-step reasoning for complex mathematical problem solving.
Input
$0.40/M
Output
$1.20/M
Context
128K
DeepSeek
Vision-language model for image understanding, OCR, and visual reasoning tasks.
Input
$0.14/M
Output
$0.28/M
Context
128K
DeepSeek
Math-specialized model achieving gold-level scores in math competitions. Based on V3.2 architecture.
Input
$0.27/M
Output
$1.10/M
Context
128K
DeepSeek
Distilled R1 reasoning into compact Qwen-based model. Exceptional at math and programming.
Input
$0.07/M
Output
$0.14/M
Context
128K
DeepSeek
R1 reasoning distilled into Llama 3 architecture. Strong reasoning at minimal compute cost.
Input
$0.07/M
Output
$0.14/M
Context
128K
Mistral AI
First Mistral reasoning model with 50% AIME-24 improvement via scalable RL. Reasoning in 8+ languages.
Input
$2.00/M
Output
$6.00/M
Context
128K
Mistral AI
Open-source reasoning model built on Small 3.1 with SFT and RL training. Efficient multilingual reasoning.
Input
$0.20/M
Output
$0.60/M
Context
128K
Mistral AI
Coding-specialized model outperforming Qwen 3 Coder Flash despite smaller size.
Input
$0.20/M
Output
$0.60/M
Context
128K
Mistral AI
Smallest Mistral model for edge computing and extremely resource-constrained deployments.
Input
$0.04/M
Output
$0.10/M
Context
128K
Mistral AI
Code model using Mamba SSM architecture for linear-time inference. Unlimited theoretical context.
Input
$0.10/M
Output
$0.30/M
Context
256K
Microsoft
Enhanced reasoning model using 1.5x more tokens for higher accuracy on complex logical tasks.
Input
$0.07/M
Output
$0.14/M
Context
32K
NVIDIA
Hybrid Mamba-Transformer MoE with 4x higher throughput than predecessor. Open weights and training data.
Input
$0.04/M
Output
$0.08/M
Context
1.0M
NVIDIA
Speed-optimized ASR model delivering 1000+ RTFx on Open ASR Leaderboard. Exceptional accuracy.
Input
$0.0040/M
Output
$0.0040/M
Cohere
Multilingual model covering 23 languages for global enterprise deployment.
Input
$0.05/M
Output
$0.15/M
Context
128K
Cohere
State-of-the-art text embedding model for semantic search and RAG applications.
Input
$0.10/M
Output
$0.10/M
Context
8K
TII
Outperforms all models under 13B on HuggingFace leaderboard. Trained on 14T tokens with innovative 1.58-bit quantized variant.
Input
$0.10/M
Output
$0.30/M
Context
32K
TII
Versatile 7B model with 30 checkpoint variants including base, instruct, and quantized.
Input
$0.07/M
Output
$0.21/M
Context
32K
Nomic AI
First MoE embedding model. Trained on 1.6B pairs across ~100 languages with top-2 expert routing.
Input
$0.01/M
Output
$0.01/M
Context
8K
Jina AI
Universal multimodal embedding handling text, images, and documents in 30+ languages.
Input
$0.02/M
Output
$0.02/M
Context
8K
Amazon
Fast, cost-effective reasoning model with built-in code interpreter and web grounding.
Input
$0.80/M
Output
$2.40/M
Context
1.0M
Amazon
Speech-to-speech model for natural real-time conversations. Supports 7 languages.
Input
$0.50/M
Output
$0.50/M
Amazon
Image generation model with fine-grained control over composition, style, and content.
Input
$0.04/M
Output
$0.04/M
Apple
On-device model optimized for Apple silicon with 2-bit quantization. Powers Siri and Apple Intelligence.
Input
Free/M
Output
Free/M
Context
4K
Reka AI
Full multimodal model handling text, image, video, and audio inputs natively.
Input
$3.00/M
Output
$9.00/M
Context
128K
Reka AI
One of the few 21B models supporting full interleaved multimodal inputs. Videos up to 5 minutes.
Input
$0.80/M
Output
$2.40/M
Context
128K
Genmo
High-performance open text-to-video model excelling in text consistency.
Input
$0.05/M
Output
$0.05/M
IBM
Enterprise-grade model with strong instruction following for business applications.
Input
$0.10/M
Output
$0.20/M
Context
128K
IBM
Compact enterprise model for edge deployment and lightweight business tasks.
Input
$0.03/M
Output
$0.06/M
Context
128K
IBM
Updated Granite with enhanced coding and tool-use capabilities for enterprise automation.
Input
$0.10/M
Output
$0.20/M
Context
128K
IBM
Small enterprise model with coding support for lightweight automation workflows.
Input
$0.03/M
Output
$0.06/M
Context
128K
OpenBMB
Efficient vision-language model rivaling GPT-4V quality at a fraction of the size.
Input
$0.10/M
Output
$0.20/M
Context
128K
Hugging Face
Compact LLM designed for on-device AI. Surprisingly capable for its tiny size.
Input
$0.01/M
Output
$0.02/M
Context
8K
Hugging Face
Tiny but functional language model for extreme resource constraints and research.
Input
$0.0050/M
Output
$0.01/M
Context
8K
Naver
Korean sovereign AI with omnimodal capabilities. Specialized for Korean language and culture.
Input
$1.00/M
Output
$3.00/M
Context
128K
Mistral AI
Mid-size Mistral model bridging the gap between 8B edge models and large frontier offerings.
Input
$0.15/M
Output
$0.45/M
Context
128K
Alibaba/Qwen
Compact math-specialized model with chain-of-thought reasoning for mathematical problem solving.
Input
$0.07/M
Output
$0.14/M
Context
128K
Microsoft
Lightweight multimodal model with vision capabilities for on-device and edge visual understanding.
Input
$0.05/M
Output
$0.10/M
Context
128K
Amazon
Amazon's video generation model producing high-quality short clips for advertising and social media.
Input
$0.04/M
Output
$0.04/M
NVIDIA
NVIDIA-tuned Llama 3.1 with reward-model-guided alignment. Excels at instruction following and helpful responses.
Input
$0.35/M
Output
$1.05/M
Context
128K
Cohere
Cohere's smallest Command model optimized for RAG, tool use, and multilingual enterprise applications.
Input
$0.04/M
Output
$0.08/M
Context
128K
Zhipu AI
Zhipu's latest generation model with improved reasoning, coding, and multilingual capabilities.
Input
$1.50/M
Output
$4.50/M
Context
128K
Experimental Gemini model with extended chain-of-thought reasoning. Transparent thinking process with strong performance on math and science.
Input
$0.15/M
Output
$0.60/M
Context
1.0M
Cohere
Cohere's open multimodal model for visual understanding across 23 languages. Strong image captioning and visual QA.
Input
$0.25/M
Output
$0.50/M
Context
128K
OpenAI
A smaller, faster, and more affordable version of GPT-4o. Great for lightweight tasks.
Input
$0.15/M
Output
$0.60/M
Context
128K
Anthropic
Anthropic's fastest and most affordable model. Great for high-volume, low-latency tasks.
Input
$0.80/M
Output
$4.00/M
Context
200K
Google's previous-gen flagship model with the longest context window in production.
Input
$1.25/M
Output
$5.00/M
Context
2.1M
Meta
Meta's largest open-source model. Competitive with frontier closed-source models.
Input
$0.80/M
Output
$0.80/M
Context
128K
Mistral AI
Mistral's flagship model with strong multilingual and code generation capabilities.
Input
$2.00/M
Output
$6.00/M
Context
128K
Mistral AI
Mistral's efficient model for everyday tasks. Fast and cost-effective.
Input
$0.10/M
Output
$0.30/M
Context
32K
xAI
xAI's large language model with real-time X (Twitter) data access and strong reasoning.
Input
$2.00/M
Output
$10.00/M
Context
131K
Cohere
Cohere's enterprise-grade model optimized for RAG, tool use, and business workflows.
Input
$2.50/M
Output
$10.00/M
Context
128K
OpenAI
A fast, affordable variant of GPT-4.1 for high-volume workloads.
Input
$0.40/M
Output
$1.60/M
Context
1.0M
Moonshot AI
Moonshot AI's reasoning-focused MoE model with chain-of-thought capabilities. 1T total params, 32B active.
Input
$0.47/M
Output
$2.00/M
Context
131K
Alibaba/Qwen
Hosted version of Qwen3.5 397B with 1M context window and adaptive thinking for complex tasks.
Input
$0.40/M
Output
$2.40/M
Context
1.0M
Alibaba/Qwen
Alibaba's large-scale reasoning model with ~1T parameters and chain-of-thought capabilities.
Input
$1.20/M
Output
$6.00/M
Context
256K
Alibaba/Qwen
Alibaba's efficient code-focused MoE model. 80B total params, 3B active, Apache 2.0 licensed.
Input
$0.12/M
Output
$0.75/M
Context
256K
OpenAI
OpenAI's fastest and cheapest model. Ideal for classification, autocomplete, and high-throughput tasks.
Input
$0.10/M
Output
$0.40/M
Context
1.0M
OpenAI
OpenAI's research preview with improved emotional intelligence and reduced hallucinations.
Input
$75.00/M
Output
$150.00/M
Context
128K
Anthropic
Upgraded Claude 3.5 Sonnet with major coding and tool-use improvements, plus computer use capability.
Input
$3.00/M
Output
$15.00/M
Context
200K
Google's open-source multimodal model. Strong performance for its size with vision capabilities.
Input
$0.10/M
Output
$0.10/M
Context
128K
Efficient open-source model from Google with multimodal capabilities at 12B parameters.
Input
$0.05/M
Output
$0.05/M
Context
128K
Ultra-efficient open-source model from Google. Runs on mobile and edge devices.
Input
$0.02/M
Output
$0.02/M
Context
128K
Google's previous-gen open-source model with strong general capabilities.
Input
$0.07/M
Output
$0.07/M
Context
8K
Efficient open-source model from Google. Great performance-to-size ratio.
Input
$0.03/M
Output
$0.03/M
Context
8K
Google's open-source code-focused model based on the Gemma architecture.
Input
$0.03/M
Output
$0.03/M
Context
8K
Meta
Meta's largest multimodal Llama model with image understanding capabilities.
Input
$0.35/M
Output
$0.40/M
Context
128K
Meta
Efficient multimodal Llama model for image + text tasks at 11B parameters.
Input
$0.06/M
Output
$0.06/M
Context
128K
Meta
Ultra-lightweight Llama model for edge deployment and mobile applications.
Input
$0.01/M
Output
$0.01/M
Context
128K
Meta
The smallest Llama model for on-device inference and constrained environments.
Input
$0.01/M
Output
$0.01/M
Context
128K
Meta
Meta's efficient open-source base model. Excellent for fine-tuning and custom deployments.
Input
$0.05/M
Output
$0.05/M
Context
128K
Meta
Meta's strong mid-range open-source model, predecessor to 3.3 with broad community support.
Input
$0.18/M
Output
$0.18/M
Context
128K
Meta
Meta's largest code-focused open-source model. Specialized for code generation and understanding.
Input
$0.18/M
Output
$0.18/M
Context
16K
Mistral AI
Mistral's first code-focused model with 32K context. Supports 80+ programming languages.
Input
$0.30/M
Output
$0.90/M
Context
32K
Mistral AI
Mistral's 12B open-source model co-developed with NVIDIA. Replaces Mistral 7B.
Input
$0.04/M
Output
$0.04/M
Context
128K
Mistral AI
Mistral's open-source multimodal model. Processes images natively alongside text.
Input
$0.10/M
Output
$0.10/M
Context
128K
Mistral AI
Mistral's flagship multimodal model. Built on Mistral Large with vision capabilities.
Input
$2.00/M
Output
$6.00/M
Context
128K
Mistral AI
Mistral's large open-source MoE model with 176B total params. Strong coding and reasoning.
Input
$0.65/M
Output
$0.65/M
Context
66K
Mistral AI
The original open-source MoE model that started the MoE trend. Fast and efficient.
Input
$0.24/M
Output
$0.24/M
Context
32K
Mistral AI
Mistral's edge-optimized model with a knowledge-dense 8B parameter design.
Input
$0.10/M
Output
$0.10/M
Context
128K
Mistral AI
The model that launched Mistral. Open-source, fast, and surprisingly capable for 7B.
Input
$0.06/M
Output
$0.06/M
Context
32K
DeepSeek
DeepSeek's open-source code-focused MoE model. Competitive with GPT-4 Turbo on coding.
Input
$0.14/M
Output
$0.28/M
Context
128K
DeepSeek
Merged general and coder capabilities from V2 into a unified model.
Input
$0.14/M
Output
$0.28/M
Context
128K
DeepSeek
R1 reasoning capabilities distilled into a compact Qwen-based 32B model.
Input
$0.12/M
Output
$0.18/M
Context
128K
Cohere
Cohere's open-weight model optimized for RAG and tool use. Strong multilingual support.
Input
$0.15/M
Output
$0.60/M
Context
128K
Cohere
Cohere's open-source multilingual model covering 23 languages with strong performance.
Input
$0.50/M
Output
$1.50/M
Context
128K