39 Tracked AI Models Across 9 Provider Ecosystems
Browse complete token specifications, input/cached/output costs, context window depths, and provider rate limit quotas across all 39 tracked models. Includes active flagships, developer reasoning tiers, and legacy reference baselines for architectural cost benchmarking.
Next generation intelligence for long-running autonomous agents
Ideal for complex agentic coding, deep domain synthesis, and enterprise work
High-performance frontier model for production coding and complex agents
Fastest, most cost-efficient model for sub-second high-volume tasks
OpenAI frontier flagship omni model for autonomous tasks and deep cross-modal reasoning
High-capability general intelligence workhorse with promotional 2026 pricing
Balanced production model delivering top-tier performance at everyday rates
Ultra-low-latency budget model engineered for high-frequency micro-tasks
Next-generation reasoning model optimized for mathematical rigor and complex software engineering
Maximum compute reasoning model for PhD-level research and formal theorem proving
High-speed reasoning model specialized for STEM, competitive code, and analytical pipelines
Google frontier Flash model with native 1M context, sub-second latency, and multimodal video support
Frontier 1.05M context Preview model with multimodal reasoning, coding depth, and tiered pricing
Ultra-economical 1M context model for automated high-volume bulk pipelines
Flagship 1M context architecture with tiered peak/off-peak pricing (Cache hit: $0.006/M peak, $0.003/M off-peak)
Meta open architecture with 327,680 tokens supported on Together AI serverless API endpoint (base architecture supports up to 1M context on dedicated clusters)
Meta flagship open-weights MoE model designed for high-reasoning multimodal dialogue
Mistral flagship reasoning and coding model with aggressive 2026 pricing ($0.50/M in)
High-speed compact model delivering low-latency inference at minimal compute cost
xAI flagship frontier model with 500k context, real-time reasoning, and prompt caching
Second-generation multimodal frontier model from xAI; superseded by Grok 4.6
High-speed compact model from xAI; superseded by current Grok 4 series
High-capacity 72B multilingual instruction model hosted on Together AI serverless endpoints
High-efficiency open-weights coding model hosted on Together AI serverless infrastructure
High-capacity multilingual enterprise RAG engine; predecessor to Cohere's Command A series
Fast enterprise RAG workhorse; predecessor to Cohere's Command A series
Specialized generative coding and fill-in-the-middle model fluent in 80+ programming languages
Predecessor 671B MoE architecture; DeepSeek routes legacy deepseek-chat requests to V4.1-Flash
High-throughput sub-second multimodal model with native 1M context at budget rates
Massive 405B dense parameter open-weights baseline; predecessor to Llama 4 family
First-generation reasoning model; succeeded by OpenAI o3 and o3-pro
High-throughput compact model for classification, extraction, and routing workloads
Fast mid-tier reasoning successor to o4-mini for STEM, code, and agentic pipelines
Generally available flagship with 1M native context and tiered >200k long-context pricing
Deep-thinking reasoning tier with 98% prefix-cache discount and 50% diurnal off-peak pricing
Open-weights balanced model with native long-context support across Together and first-party serving
MoE flagship balancing frontier quality with framework-native serving costs
Cohere flagship with enterprise RAG tooling and automatic generation-time caching
High-efficiency xAI model for classification and fast agent loops