LLM API pricing comparison
Published API prices per million tokens for 32 models across 8 providers. Sort by any column, or open a model for worked examples of what it costs at real volumes.
Prices verified 4 Aug 2026Model your own workload →
| Provider | Cached in | Batch | Verified | |||||
|---|---|---|---|---|---|---|---|---|
| Llama 3.1 8B (Groq) | Groq | $0.050 | $0.080 | — | 50% off | $0.058 | 128K | 4 Aug 2026 |
| Amazon Nova Micro | Bedrock | $0.035 | $0.140 | — | 50% off | $0.061 | 128K | 4 Aug 2026 |
| Amazon Nova Lite | Bedrock | $0.060 | $0.240 | — | 50% off | $0.105 | 300K | 4 Aug 2026 |
| GPT-5 nano | OpenAI | $0.050 | $0.400 | $0.00500 | 50% off | $0.138 | 400K | 4 Aug 2026 |
| GPT-4.1 nano | OpenAI | $0.100 | $0.400 | $0.025 | 50% off | $0.175 | 1.0M | 4 Aug 2026 |
| Gemini 2.5 Flash-Lite | $0.100 | $0.400 | $0.010 | 50% off | $0.175 | 1.0M | 4 Aug 2026 | |
| Gemini 2.0 Flash | $0.100 | $0.400 | $0.025 | 50% off | $0.175 | 1.0M | 4 Aug 2026 | |
| DeepSeek V4 Flash | DeepSeek | $0.140 | $0.280 | $0.00280 | — | $0.175 | 1.0M | 4 Aug 2026 |
| GPT-4o mini | OpenAI | $0.150 | $0.600 | $0.075 | 50% off | $0.262 | 128K | 4 Aug 2026 |
| Mistral Small 4 | Mistral | $0.150 | $0.600 | $0.015 | 50% off | $0.262 | 131K | 4 Aug 2026 |
| DeepSeek V4 Pro | DeepSeek | $0.435 | $0.870 | $0.00363 | — | $0.544 | 1.0M | 4 Aug 2026 |
| Llama 3.3 70B (Groq) | Groq | $0.590 | $0.790 | — | 50% off | $0.640 | 128K | 4 Aug 2026 |
| GPT-5 mini | OpenAI | $0.250 | $2.00 | $0.025 | 50% off | $0.688 | 400K | 4 Aug 2026 |
| GPT-4.1 mini | OpenAI | $0.400 | $1.60 | $0.100 | 50% off | $0.700 | 1.0M | 4 Aug 2026 |
| Mistral Large 3 | Mistral | $0.500 | $1.50 | $0.050 | 50% off | $0.750 | 131K | 4 Aug 2026 |
| Gemini 2.5 Flash | $0.300 | $2.50 | $0.030 | 50% off | $0.850 | 1.0M | 4 Aug 2026 | |
| Amazon Nova Pro | Bedrock | $0.800 | $3.20 | — | 50% off | $1.40 | 300K | 4 Aug 2026 |
| Grok 4.3 | xAI | $1.25 | $2.50 | $0.200 | — | $1.56 | 1.0M | 4 Aug 2026 |
| Claude Haiku 3.5 | Anthropic | $0.800 | $4.00 | $0.080 | 50% off | $1.60 | 200K | 4 Aug 2026 |
| o4-mini | OpenAI | $1.10 | $4.40 | $0.275 | 50% off | $1.93 | 200K | 4 Aug 2026 |
| Claude Haiku 4.5 | Anthropic | $1.00 | $5.00 | $0.100 | 50% off | $2.00 | 200K | 4 Aug 2026 |
| Grok 4.5 | xAI | $2.00 | $6.00 | $0.300 | — | $3.00 | 500K | 4 Aug 2026 |
| GPT-5 | OpenAI | $1.25 | $10.00 | $0.125 | 50% off | $3.44 | 400K | 4 Aug 2026 |
| Gemini 2.5 Pro | $1.25 | $10.00 | $0.125 | 50% off | $3.44 | 1.0M | 4 Aug 2026 | |
| GPT-4.1 | OpenAI | $2.00 | $8.00 | $0.500 | 50% off | $3.50 | 1.0M | 4 Aug 2026 |
| o3 | OpenAI | $2.00 | $8.00 | $0.500 | 50% off | $3.50 | 200K | 4 Aug 2026 |
| Claude Sonnet 5 | Anthropic | $2.00 | $10.00 | $0.200 | 50% off | $4.00 | 1.0M | 4 Aug 2026 |
| GPT-4o | OpenAI | $2.50 | $10.00 | $1.25 | 50% off | $4.38 | 128K | 4 Aug 2026 |
| Amazon Nova Premier | Bedrock | $2.50 | $12.50 | — | 50% off | $5.00 | 1.0M | 4 Aug 2026 |
| Claude Sonnet 4.5 | Anthropic | $3.00 | $15.00 | $0.300 | 50% off | $6.00 | 200K | 4 Aug 2026 |
| Claude Opus 5 | Anthropic | $5.00 | $25.00 | $0.500 | 50% off | $10.00 | 1.0M | 4 Aug 2026 |
| Claude Opus 4.1 | Anthropic | $15.00 | $75.00 | $1.50 | 50% off | $30.00 | 200K | 4 Aug 2026 |
“Blended” is the effective $/1M at a 3:1 input-to-output ratio — a single number for comparing models whose input and output prices differ in shape. Your own ratio is what matters; use the calculator for that.
Popular comparisons
The matchups people actually have to decide between.
Claude Sonnet 5 vs GPT-5$4.00 vs $3.44 blendedClaude Opus 5 vs GPT-5$10.00 vs $3.44 blendedGPT-5 vs Gemini 2.5 Pro$3.44 vs $3.44 blendedGrok 4.5 vs GPT-5$3.00 vs $3.44 blendedGPT-5 vs GPT-5 mini$3.44 vs $0.688 blendedClaude Opus 5 vs Claude Sonnet 5$10.00 vs $4.00 blendedGemini 2.5 Pro vs Gemini 2.5 Flash$3.44 vs $0.850 blendedClaude Sonnet 4.5 vs Claude Haiku 4.5$6.00 vs $2.00 blendedClaude Sonnet 4.5 vs GPT-4.1$6.00 vs $3.50 blendedGPT-5 mini vs Gemini 2.5 Flash$0.688 vs $0.850 blendedClaude Haiku 4.5 vs GPT-5 nano$2.00 vs $0.138 blendedClaude Haiku 4.5 vs Gemini 2.5 Flash-Lite$2.00 vs $0.175 blendedGemini 2.5 Flash vs GPT-4o mini$0.850 vs $0.262 blendedo3 vs o4-mini$3.50 vs $1.93 blendedDeepSeek V4 Pro vs o4-mini$0.544 vs $1.93 blendedDeepSeek V4 Flash vs GPT-4o mini$0.175 vs $0.262 blendedLlama 3.3 70B (Groq) vs GPT-4o mini$0.640 vs $0.262 blendedMistral Large 3 vs GPT-4.1$0.750 vs $3.50 blendedAmazon Nova Pro vs Claude Sonnet 4.5$1.40 vs $6.00 blended
By provider
Anthropic Claude
DeepSeek
Mistral AI
Cheapest blended rate in the catalog: Llama 3.1 8B (Groq) at $0.058 per 1M tokens. Prices are vendor list prices and exclude committed-use or enterprise discounts.
Knowing the rate card is the easy half.
Finitizer TokenOps shows which teams, features, and prompts are actually consuming your token budget — with per-model attribution, budgets, and anomaly alerts across Anthropic, OpenAI, and Google.