Model Costs
Live token pricing across all major AI providers · Updated daily
Provider:
Tier:
| Provider | Model | Context | Input / 1M tokens ↑ | Output / 1M tokens | Tier | Notes |
|---|---|---|---|---|---|---|
| Meta/Groq | Llama 3.1 8B Instant | 128K | $0.05 | $0.08 | budget | Ultra-low cost, ultra-fast inference for lightweight tasks. |
| Gemini 3.5 Flash | 1M | $0.07 | $0.30 | budget | Fast, low-cost multimodal model for high-volume applications. | |
| Gemini 2.5 Flash | 1M | $0.07 | $0.30 | budget | Competitive budget-tier multimodal model with fast inference. | |
| Mistral | Mistral Small | 32K | $0.15 | $0.45 | budget | Budget-friendly option with caching and batch discounts. |
| Meta/Together AI | Llama 3.1 8B | 128K | $0.18 | $0.18 | budget | Lightweight model with low cost and caching support. |
| Mistral | Mistral Codestral | 32K | $0.25 | $0.75 | budget | Specialized code generation model with caching support. |
| DeepSeek | DeepSeek V3 | 64K | $0.27 | $1.10 | budget | Competitive pricing with strong performance across tasks. |
| Mistral | Mistral Nemo | 128K | $0.30 | $0.90 | budget | Efficient model with extended context and batch optimization. |
| Perplexity | Sonar | 128K | $0.50 | $1.50 | budget | Budget variant with search integration capabilities. |
| DeepSeek | DeepSeek R1 | 64K | $0.55 | $2.19 | standard | Reasoning-focused variant with enhanced problem-solving. |
| Meta/Groq | Llama 3.3 70B Versatile | 128K | $0.59 | $0.79 | standard | High-speed inference on Groq hardware with strong versatility. |
| Meta/Groq | Llama 3.1 405B | 128K | $0.59 | $0.79 | standard | Largest Llama variant optimized for speed and cost on Groq. |
| Meta/Groq | Llama 3.1 70B | 128K | $0.59 | $0.79 | standard | Mid-size Llama model with excellent speed-to-cost ratio. |
| Meta/Together AI | Llama 3.1 70B | 128K | $0.60 | $0.60 | standard | Mid-tier Llama with flexible deployment and caching options. |
| Mistral | Mistral Pixtral | 128K | $0.60 | $1.80 | standard | Vision-capable model for multimodal applications. |
| OpenAI | GPT-5.4-mini | 128K | $0.75 | $4.50 | budget | Affordable mini variant for cost-sensitive applications. |
| Meta/Together AI | Llama 3.1 405B | 128K | $0.88 | $0.88 | standard | Largest model with caching support on Together AI platform. |
| xAI | Grok-build-0.1 | 256K | $1.00 | $2.00 | budget | Fast coding and agentic-focused variant with budget-tier pricing. |
| Meta/Together AI | Llama 3.3 70B | 128K | $1.04 | $1.04 | standard | Flexible serverless and dedicated deployment options available. |
| Gemini 2.5 Pro | 1M | $1.25 | $5.00 | standard | Advanced reasoning and multimodal capabilities with context-length pricing tiers. | |
| xAI | Grok-4.3 | 1M | $1.25 | $2.50 | standard | Strong reasoning and tool-calling with caching options available. |
| Gemini 3 Pro | 1M | $1.50 | $6.00 | standard | Premium reasoning model with enhanced multimodal support. | |
| Mistral | Mistral Large | 128K | $2.00 | $6.00 | standard | Current flagship with 50% batch processing discount available. |
| OpenAI | GPT-5.4 | 128K | $2.50 | $15.00 | standard | Professional variant balancing cost and performance for coding and work tasks. |
| Anthropic | Claude 4.6 Sonnet | 1M | $3.00 | $15.00 | standard | Strong agentic and coding performance with 1M context window in beta. |
| Cohere | Command R+ | 128K | $3.00 | $15.00 | standard | Enterprise-grade model with strong reasoning capabilities. |
| Perplexity | Sonar Pro | 200K | $3.00 | $15.00 | standard | Search-integrated model with extended context window. |
| OpenAI | GPT-5.5 | 128K+ | $5.00 | $30.00 | premium | Flagship reasoning and coding model with caching discounts available. |
| Anthropic | Claude 4.8 Opus | 200K | $5.00 | $25.00 | premium | Most capable Claude model for complex reasoning and agentic tasks. |
| OpenAI | GPT-5.5 Long | 128K+ | $10.00 | $45.00 | premium | Extended context variant of GPT-5.5 for longer documents. |
Prices in USD per 1M tokens · Updated daily · Always verify against official provider documentation