{
  "name": "LLM Cost Lab — LLM API pricing dataset",
  "description": "Standard public list prices for large language model APIs, in USD per 1,000,000 tokens.",
  "source": "https://llmcostlab.com",
  "license": "CC BY 4.0 — free to reuse with attribution.",
  "lastVerified": "2026-08-16",
  "currency": "USD",
  "unit": "per 1,000,000 tokens",
  "modelCount": 60,
  "blendedRateDefinition": "blendedPerM assumes 3 input tokens per 1 output token, which is closer to typical chat and RAG traffic than either rate alone.",
  "models": [
    {
      "id": "gpt-5-6-sol",
      "name": "GPT-5.6 Sol",
      "provider": "OpenAI",
      "inputPerMillionTokens": 5,
      "outputPerMillionTokens": 30,
      "cachedInputPerMillionTokens": 0.5,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 11.25,
      "notes": "Flagship reasoning tier. Long-context prompts may incur a surcharge: verify on the official pricing page for very large inputs.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-6-sol/"
    },
    {
      "id": "gpt-5-6-terra",
      "name": "GPT-5.6 Terra",
      "provider": "OpenAI",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 12,
      "cachedInputPerMillionTokens": 0.2,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 4.5,
      "notes": "Mid tier. Repriced on July 30, 2026, roughly 20% below its launch rate of $2.50 / $15.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-6-terra/"
    },
    {
      "id": "gpt-5-6-luna",
      "name": "GPT-5.6 Luna",
      "provider": "OpenAI",
      "inputPerMillionTokens": 0.2,
      "outputPerMillionTokens": 1.2,
      "cachedInputPerMillionTokens": 0.02,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.45,
      "notes": "Smallest current-gen tier. Cut roughly 80% on July 30, 2026, down from $1 / $6 at launch.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-6-luna/"
    },
    {
      "id": "gpt-5-5",
      "name": "GPT-5.5",
      "provider": "OpenAI",
      "inputPerMillionTokens": 5,
      "outputPerMillionTokens": 30,
      "cachedInputPerMillionTokens": 0.5,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 11.25,
      "notes": "Previous-generation flagship, still listed at the same rate as GPT-5.6 Sol.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-5/"
    },
    {
      "id": "gpt-5-5-pro",
      "name": "GPT-5.5 Pro",
      "provider": "OpenAI",
      "inputPerMillionTokens": 30,
      "outputPerMillionTokens": 180,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 67.5,
      "notes": "Extended-compute tier with no cached-input discount. The most expensive model tracked on this site.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-5-pro/"
    },
    {
      "id": "gpt-5-4",
      "name": "GPT-5.4",
      "provider": "OpenAI",
      "inputPerMillionTokens": 2.5,
      "outputPerMillionTokens": 15,
      "cachedInputPerMillionTokens": 0.25,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 5.625,
      "notes": "Prior mid tier, now priced above GPT-5.6 Terra: check whether the newer model is a straight upgrade for your workload.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-4/"
    },
    {
      "id": "gpt-5-4-mini",
      "name": "GPT-5.4 mini",
      "provider": "OpenAI",
      "inputPerMillionTokens": 0.75,
      "outputPerMillionTokens": 4.5,
      "cachedInputPerMillionTokens": 0.075,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.6875,
      "notes": null,
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-4-mini/"
    },
    {
      "id": "gpt-5-4-nano",
      "name": "GPT-5.4 nano",
      "provider": "OpenAI",
      "inputPerMillionTokens": 0.2,
      "outputPerMillionTokens": 1.25,
      "cachedInputPerMillionTokens": 0.02,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.4625,
      "notes": null,
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-5-4-nano/"
    },
    {
      "id": "gpt-4-1-nano",
      "name": "GPT-4.1 nano",
      "provider": "OpenAI",
      "inputPerMillionTokens": 0.1,
      "outputPerMillionTokens": 0.4,
      "cachedInputPerMillionTokens": 0.025,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.175,
      "notes": "Cheapest OpenAI model by input price. Older generation, kept available for high-volume simple tasks.",
      "providerPricingPage": "https://platform.openai.com/docs/pricing",
      "page": "https://llmcostlab.com/models/gpt-4-1-nano/"
    },
    {
      "id": "claude-fable-5",
      "name": "Claude Fable 5",
      "provider": "Anthropic",
      "inputPerMillionTokens": 10,
      "outputPerMillionTokens": 50,
      "cachedInputPerMillionTokens": 1,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 20,
      "notes": "Anthropic's most capable widely released model, for demanding reasoning and long-horizon agent work.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-fable-5/"
    },
    {
      "id": "claude-opus-5",
      "name": "Claude Opus 5",
      "provider": "Anthropic",
      "inputPerMillionTokens": 5,
      "outputPerMillionTokens": 25,
      "cachedInputPerMillionTokens": 0.5,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 10,
      "notes": "Current Opus tier: complex agentic coding and enterprise work at half Fable 5's rate.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-opus-5/"
    },
    {
      "id": "claude-opus-4-8",
      "name": "Claude Opus 4.8",
      "provider": "Anthropic",
      "inputPerMillionTokens": 5,
      "outputPerMillionTokens": 25,
      "cachedInputPerMillionTokens": 0.5,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 10,
      "notes": "Prior-generation Opus, priced identically to Opus 5. Useful as a pinned fallback target.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-opus-4-8/"
    },
    {
      "id": "claude-sonnet-5",
      "name": "Claude Sonnet 5",
      "provider": "Anthropic",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 10,
      "cachedInputPerMillionTokens": 0.2,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 4,
      "notes": "Introductory pricing through August 31, 2026. The standard rate is $3 / $15, so budget for the step-up.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-sonnet-5/"
    },
    {
      "id": "claude-sonnet-4-6",
      "name": "Claude Sonnet 4.6",
      "provider": "Anthropic",
      "inputPerMillionTokens": 3,
      "outputPerMillionTokens": 15,
      "cachedInputPerMillionTokens": 0.3,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 6,
      "notes": "Previous-generation Sonnet at the standard Sonnet rate.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-sonnet-4-6/"
    },
    {
      "id": "claude-haiku-4-5",
      "name": "Claude Haiku 4.5",
      "provider": "Anthropic",
      "inputPerMillionTokens": 1,
      "outputPerMillionTokens": 5,
      "cachedInputPerMillionTokens": 0.1,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 2,
      "notes": "Fastest, cheapest current Claude model.",
      "providerPricingPage": "https://www.anthropic.com/pricing",
      "page": "https://llmcostlab.com/models/claude-haiku-4-5/"
    },
    {
      "id": "gemini-3-7-flash",
      "name": "Gemini 3.7 Flash",
      "provider": "Google",
      "inputPerMillionTokens": 0.75,
      "outputPerMillionTokens": 3.75,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 1.5,
      "notes": "Released August 13, 2026 at introductory pricing. Both rates are scheduled to double on January 1, 2027.",
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-7-flash/"
    },
    {
      "id": "gemini-3-6-flash",
      "name": "Gemini 3.6 Flash",
      "provider": "Google",
      "inputPerMillionTokens": 1.5,
      "outputPerMillionTokens": 7.5,
      "cachedInputPerMillionTokens": 0.15,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 3,
      "notes": null,
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-6-flash/"
    },
    {
      "id": "gemini-3-5-flash",
      "name": "Gemini 3.5 Flash",
      "provider": "Google",
      "inputPerMillionTokens": 1.5,
      "outputPerMillionTokens": 9,
      "cachedInputPerMillionTokens": 0.15,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 3.375,
      "notes": null,
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-5-flash/"
    },
    {
      "id": "gemini-3-1-pro",
      "name": "Gemini 3.1 Pro",
      "provider": "Google",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 12,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 4.5,
      "notes": "Rate applies up to 200K-token prompts; above that, input/output rise to roughly $4 / $18 per 1M tokens.",
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-1-pro/"
    },
    {
      "id": "gemini-3-pro",
      "name": "Gemini 3 Pro",
      "provider": "Google",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 12,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 4.5,
      "notes": "Prior Pro release, still listed at the same rate as Gemini 3.1 Pro.",
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-pro/"
    },
    {
      "id": "gemini-3-5-flash-lite",
      "name": "Gemini 3.5 Flash-Lite",
      "provider": "Google",
      "inputPerMillionTokens": 0.3,
      "outputPerMillionTokens": 2.5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 0.85,
      "notes": null,
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-3-5-flash-lite/"
    },
    {
      "id": "gemini-2-5-flash-lite",
      "name": "Gemini 2.5 Flash-Lite",
      "provider": "Google",
      "inputPerMillionTokens": 0.1,
      "outputPerMillionTokens": 0.4,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 0.175,
      "notes": "Cheapest current Gemini tier.",
      "providerPricingPage": "https://ai.google.dev/pricing",
      "page": "https://llmcostlab.com/models/gemini-2-5-flash-lite/"
    },
    {
      "id": "mistral-large-3",
      "name": "Mistral Large 3",
      "provider": "Mistral",
      "inputPerMillionTokens": 0.5,
      "outputPerMillionTokens": 1.5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.75,
      "notes": "Flagship general-purpose model, and unusually cheap for a flagship tier.",
      "providerPricingPage": "https://mistral.ai/pricing",
      "page": "https://llmcostlab.com/models/mistral-large-3/"
    },
    {
      "id": "mistral-medium-3-5",
      "name": "Mistral Medium 3.5",
      "provider": "Mistral",
      "inputPerMillionTokens": 1.5,
      "outputPerMillionTokens": 7.5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 3,
      "notes": "Priced above Mistral Large 3 on both input and output: compare on your own workload before assuming the tier names map to cost.",
      "providerPricingPage": "https://mistral.ai/pricing",
      "page": "https://llmcostlab.com/models/mistral-medium-3-5/"
    },
    {
      "id": "magistral-medium",
      "name": "Magistral Medium",
      "provider": "Mistral",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 2.75,
      "notes": "Reasoning-focused tier.",
      "providerPricingPage": "https://mistral.ai/pricing",
      "page": "https://llmcostlab.com/models/magistral-medium/"
    },
    {
      "id": "mistral-small-4",
      "name": "Mistral Small 4",
      "provider": "Mistral",
      "inputPerMillionTokens": 0.15,
      "outputPerMillionTokens": 0.6,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.2625,
      "notes": null,
      "providerPricingPage": "https://mistral.ai/pricing",
      "page": "https://llmcostlab.com/models/mistral-small-4/"
    },
    {
      "id": "grok-4-6",
      "name": "Grok 4.6",
      "provider": "xAI",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 6,
      "cachedInputPerMillionTokens": 0.5,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 3,
      "notes": "Flagship xAI model, released August 12, 2026. Prompts at or above 200K tokens bill the whole request at $4 input / $12 output.",
      "providerPricingPage": "https://x.ai/api",
      "page": "https://llmcostlab.com/models/grok-4-6/"
    },
    {
      "id": "grok-4-5",
      "name": "Grok 4.5",
      "provider": "xAI",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 6,
      "cachedInputPerMillionTokens": 0.3,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 3,
      "notes": "Same standard rates as Grok 4.6 with a cheaper cached-input rate, which can favour it for cache-heavy workloads.",
      "providerPricingPage": "https://x.ai/api",
      "page": "https://llmcostlab.com/models/grok-4-5/"
    },
    {
      "id": "grok-4-3",
      "name": "Grok 4.3",
      "provider": "xAI",
      "inputPerMillionTokens": 1.25,
      "outputPerMillionTokens": 2.5,
      "cachedInputPerMillionTokens": 0.2,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.5625,
      "notes": null,
      "providerPricingPage": "https://x.ai/api",
      "page": "https://llmcostlab.com/models/grok-4-3/"
    },
    {
      "id": "grok-build-0-1",
      "name": "Grok Build 0.1",
      "provider": "xAI",
      "inputPerMillionTokens": 1,
      "outputPerMillionTokens": 2,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.25,
      "notes": "Coding-focused tier, priced below the general Grok models.",
      "providerPricingPage": "https://x.ai/api",
      "page": "https://llmcostlab.com/models/grok-build-0-1/"
    },
    {
      "id": "grok-4-1-fast",
      "name": "Grok 4.1 Fast",
      "provider": "xAI",
      "inputPerMillionTokens": 0.2,
      "outputPerMillionTokens": 0.5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.275,
      "notes": "Cheapest current xAI tier, built for low-latency use.",
      "providerPricingPage": "https://x.ai/api",
      "page": "https://llmcostlab.com/models/grok-4-1-fast/"
    },
    {
      "id": "deepseek-v4-pro",
      "name": "DeepSeek V4 Pro",
      "provider": "DeepSeek",
      "inputPerMillionTokens": 0.435,
      "outputPerMillionTokens": 0.87,
      "cachedInputPerMillionTokens": 0.003625,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.54375,
      "notes": "Cache-hit input costs roughly $0.0036/1M, a ~99% discount over a cache miss. DeepSeek moved to a peak/off-peak schedule on August 16, 2026, so confirm the rate for your traffic window.",
      "providerPricingPage": "https://api-docs.deepseek.com/quick_start/pricing",
      "page": "https://llmcostlab.com/models/deepseek-v4-pro/"
    },
    {
      "id": "deepseek-v4-flash",
      "name": "DeepSeek V4 Flash",
      "provider": "DeepSeek",
      "inputPerMillionTokens": 0.14,
      "outputPerMillionTokens": 0.28,
      "cachedInputPerMillionTokens": 0.0028,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.175,
      "notes": "One of the cheapest frontier-adjacent models tracked on this site. Also subject to the August 2026 peak/off-peak schedule.",
      "providerPricingPage": "https://api-docs.deepseek.com/quick_start/pricing",
      "page": "https://llmcostlab.com/models/deepseek-v4-flash/"
    },
    {
      "id": "nova-premier",
      "name": "Nova Premier 1.0",
      "provider": "Amazon",
      "inputPerMillionTokens": 2.5,
      "outputPerMillionTokens": 12.5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 5,
      "notes": "Most capable current Nova tier.",
      "providerPricingPage": "https://aws.amazon.com/bedrock/pricing/",
      "page": "https://llmcostlab.com/models/nova-premier/"
    },
    {
      "id": "nova-pro",
      "name": "Nova Pro 1.0",
      "provider": "Amazon",
      "inputPerMillionTokens": 0.8,
      "outputPerMillionTokens": 3.2,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.4,
      "notes": null,
      "providerPricingPage": "https://aws.amazon.com/bedrock/pricing/",
      "page": "https://llmcostlab.com/models/nova-pro/"
    },
    {
      "id": "nova-lite",
      "name": "Nova Lite 1.0",
      "provider": "Amazon",
      "inputPerMillionTokens": 0.06,
      "outputPerMillionTokens": 0.24,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.105,
      "notes": null,
      "providerPricingPage": "https://aws.amazon.com/bedrock/pricing/",
      "page": "https://llmcostlab.com/models/nova-lite/"
    },
    {
      "id": "nova-micro",
      "name": "Nova Micro",
      "provider": "Amazon",
      "inputPerMillionTokens": 0.035,
      "outputPerMillionTokens": 0.14,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.06125,
      "notes": "Text-only. The cheapest model tracked from any US provider on this site.",
      "providerPricingPage": "https://aws.amazon.com/bedrock/pricing/",
      "page": "https://llmcostlab.com/models/nova-micro/"
    },
    {
      "id": "command-a",
      "name": "Command A",
      "provider": "Cohere",
      "inputPerMillionTokens": 2.5,
      "outputPerMillionTokens": 10,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 4.375,
      "notes": "Cohere's flagship enterprise model, with a 256K-token context window.",
      "providerPricingPage": "https://cohere.com/pricing",
      "page": "https://llmcostlab.com/models/command-a/"
    },
    {
      "id": "command-r-plus",
      "name": "Command R+",
      "provider": "Cohere",
      "inputPerMillionTokens": 2.5,
      "outputPerMillionTokens": 10,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 4.375,
      "notes": "Built for enterprise RAG and tool-use workloads.",
      "providerPricingPage": "https://cohere.com/pricing",
      "page": "https://llmcostlab.com/models/command-r-plus/"
    },
    {
      "id": "command-r7b",
      "name": "Command R7B",
      "provider": "Cohere",
      "inputPerMillionTokens": 0.0375,
      "outputPerMillionTokens": 0.15,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.065625,
      "notes": "One of the cheapest models tracked on this site, just behind Amazon Nova Micro.",
      "providerPricingPage": "https://cohere.com/pricing",
      "page": "https://llmcostlab.com/models/command-r7b/"
    },
    {
      "id": "llama-3-3-70b-versatile",
      "name": "Llama 3.3 70B Versatile (via Groq)",
      "provider": "Groq",
      "inputPerMillionTokens": 0.59,
      "outputPerMillionTokens": 0.79,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 0.64,
      "notes": "Open-weight Llama model served on Groq's LPU hardware for very low latency. Batch and prompt caching each cut rates roughly 50%, and can stack.",
      "providerPricingPage": "https://groq.com/pricing",
      "page": "https://llmcostlab.com/models/llama-3-3-70b-versatile/"
    },
    {
      "id": "llama-3-1-8b-instant",
      "name": "Llama 3.1 8B Instant (via Groq)",
      "provider": "Groq",
      "inputPerMillionTokens": 0.05,
      "outputPerMillionTokens": 0.08,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 0.0575,
      "notes": "Smallest current Groq tier, built for high-throughput, low-latency workloads.",
      "providerPricingPage": "https://groq.com/pricing",
      "page": "https://llmcostlab.com/models/llama-3-1-8b-instant/"
    },
    {
      "id": "kimi-k3",
      "name": "Kimi K3",
      "provider": "Moonshot AI",
      "inputPerMillionTokens": 3,
      "outputPerMillionTokens": 15,
      "cachedInputPerMillionTokens": 0.3,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 6,
      "notes": "Flagship Moonshot AI model, released July 16, 2026.",
      "providerPricingPage": "https://platform.moonshot.ai/docs/pricing",
      "page": "https://llmcostlab.com/models/kimi-k3/"
    },
    {
      "id": "kimi-k2-6",
      "name": "Kimi K2.6",
      "provider": "Moonshot AI",
      "inputPerMillionTokens": 0.95,
      "outputPerMillionTokens": 4,
      "cachedInputPerMillionTokens": 0.16,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.7125,
      "notes": null,
      "providerPricingPage": "https://platform.moonshot.ai/docs/pricing",
      "page": "https://llmcostlab.com/models/kimi-k2-6/"
    },
    {
      "id": "kimi-k2-5",
      "name": "Kimi K2.5",
      "provider": "Moonshot AI",
      "inputPerMillionTokens": 0.6,
      "outputPerMillionTokens": 3,
      "cachedInputPerMillionTokens": 0.1,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.2,
      "notes": "Cheapest current Kimi tier.",
      "providerPricingPage": "https://platform.moonshot.ai/docs/pricing",
      "page": "https://llmcostlab.com/models/kimi-k2-5/"
    },
    {
      "id": "glm-5-2",
      "name": "GLM-5.2",
      "provider": "Zhipu AI",
      "inputPerMillionTokens": 1.4,
      "outputPerMillionTokens": 4.4,
      "cachedInputPerMillionTokens": 0.26,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 2.15,
      "notes": "Flagship GLM model on the official API, and one of the cheapest flagship tiers tracked here.",
      "providerPricingPage": "https://docs.z.ai/guides/overview/pricing",
      "page": "https://llmcostlab.com/models/glm-5-2/"
    },
    {
      "id": "glm-5",
      "name": "GLM-5",
      "provider": "Zhipu AI",
      "inputPerMillionTokens": 0.6,
      "outputPerMillionTokens": 1.92,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.93,
      "notes": "Prior GLM-5 release, cheaper than GLM-5.2 on both input and output.",
      "providerPricingPage": "https://docs.z.ai/guides/overview/pricing",
      "page": "https://llmcostlab.com/models/glm-5/"
    },
    {
      "id": "glm-4-6",
      "name": "GLM-4.6",
      "provider": "Zhipu AI",
      "inputPerMillionTokens": 0.6,
      "outputPerMillionTokens": 2.2,
      "cachedInputPerMillionTokens": 0.11,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1,
      "notes": "Prior-generation flagship, kept available at a lower price than GLM-5.2.",
      "providerPricingPage": "https://docs.z.ai/guides/overview/pricing",
      "page": "https://llmcostlab.com/models/glm-4-6/"
    },
    {
      "id": "qwen-3-8-max",
      "name": "Qwen3.8 Max",
      "provider": "Alibaba",
      "inputPerMillionTokens": 2,
      "outputPerMillionTokens": 6,
      "cachedInputPerMillionTokens": 0.25,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 3,
      "notes": "Alibaba's flagship, released August 3, 2026. Rates shown are the International (Singapore) endpoint; the Mainland China endpoint runs 60-70% cheaper.",
      "providerPricingPage": "https://www.alibabacloud.com/help/en/model-studio/models",
      "page": "https://llmcostlab.com/models/qwen-3-8-max/"
    },
    {
      "id": "qwen-3-7-max",
      "name": "Qwen3.7 Max",
      "provider": "Alibaba",
      "inputPerMillionTokens": 1.25,
      "outputPerMillionTokens": 3.75,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 1.875,
      "notes": "Previous flagship, still listed and materially cheaper than Qwen3.8 Max.",
      "providerPricingPage": "https://www.alibabacloud.com/help/en/model-studio/models",
      "page": "https://llmcostlab.com/models/qwen-3-7-max/"
    },
    {
      "id": "qwen-3-7-flash",
      "name": "Qwen3.7 Flash",
      "provider": "Alibaba",
      "inputPerMillionTokens": 0.03,
      "outputPerMillionTokens": 0.13,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": 50,
      "blendedPerMillionTokens": 0.055,
      "notes": "The cheapest model tracked anywhere on this site, on input price.",
      "providerPricingPage": "https://www.alibabacloud.com/help/en/model-studio/models",
      "page": "https://llmcostlab.com/models/qwen-3-7-flash/"
    },
    {
      "id": "llama-4-maverick",
      "name": "Llama 4 Maverick",
      "provider": "Meta",
      "inputPerMillionTokens": 0.2,
      "outputPerMillionTokens": 0.6,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.3,
      "notes": "Meta does not run a first-party paid API. These are representative rates from hosted providers; the exact figure varies by host, so check yours.",
      "providerPricingPage": "https://www.llama.com/",
      "page": "https://llmcostlab.com/models/llama-4-maverick/"
    },
    {
      "id": "llama-4-scout",
      "name": "Llama 4 Scout",
      "provider": "Meta",
      "inputPerMillionTokens": 0.08,
      "outputPerMillionTokens": 0.3,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.135,
      "notes": "Smaller open-weight Llama 4 tier. Hosted rates vary by provider; self-hosting shifts the cost to GPU time instead.",
      "providerPricingPage": "https://www.llama.com/",
      "page": "https://llmcostlab.com/models/llama-4-scout/"
    },
    {
      "id": "sonar-pro",
      "name": "Sonar Pro",
      "provider": "Perplexity",
      "inputPerMillionTokens": 3,
      "outputPerMillionTokens": 15,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 6,
      "notes": "Search-grounded model. Token rates are only part of the bill: Sonar models also charge a per-request search fee of roughly $5-$14 per 1,000 requests depending on search depth.",
      "providerPricingPage": "https://docs.perplexity.ai/getting-started/pricing",
      "page": "https://llmcostlab.com/models/sonar-pro/"
    },
    {
      "id": "sonar-huge-online",
      "name": "Sonar Huge Online",
      "provider": "Perplexity",
      "inputPerMillionTokens": 5,
      "outputPerMillionTokens": 5,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 5,
      "notes": "Flat input/output rate. Per-request search fees apply on top.",
      "providerPricingPage": "https://docs.perplexity.ai/getting-started/pricing",
      "page": "https://llmcostlab.com/models/sonar-huge-online/"
    },
    {
      "id": "sonar",
      "name": "Sonar",
      "provider": "Perplexity",
      "inputPerMillionTokens": 1,
      "outputPerMillionTokens": 1,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1,
      "notes": "Flat $1 in / $1 out. Per-request search fees apply on top.",
      "providerPricingPage": "https://docs.perplexity.ai/getting-started/pricing",
      "page": "https://llmcostlab.com/models/sonar/"
    },
    {
      "id": "sonar-small-online",
      "name": "Sonar Small Online",
      "provider": "Perplexity",
      "inputPerMillionTokens": 0.2,
      "outputPerMillionTokens": 0.2,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.2,
      "notes": "Cheapest Sonar tier. Per-request search fees apply on top.",
      "providerPricingPage": "https://docs.perplexity.ai/getting-started/pricing",
      "page": "https://llmcostlab.com/models/sonar-small-online/"
    },
    {
      "id": "doubao-seed-2-1-pro",
      "name": "Doubao Seed 2.1 Pro",
      "provider": "ByteDance",
      "inputPerMillionTokens": 0.85,
      "outputPerMillionTokens": 4.23,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.695,
      "notes": "ByteDance quotes this model in CNY (¥6 in / ¥30 out per 1M). Figures here are converted at roughly ¥7.1 to the dollar and will drift with the exchange rate.",
      "providerPricingPage": "https://www.volcengine.com/docs/82379",
      "page": "https://llmcostlab.com/models/doubao-seed-2-1-pro/"
    },
    {
      "id": "ernie-5-1",
      "name": "ERNIE 5.1",
      "provider": "Baidu",
      "inputPerMillionTokens": 0.56,
      "outputPerMillionTokens": 2.54,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 1.055,
      "notes": "Quoted by Baidu in CNY (¥4 in / ¥18 out per 1M). Converted at roughly ¥7.1 to the dollar; treat as approximate.",
      "providerPricingPage": "https://cloud.baidu.com/product/wenxinworkshop",
      "page": "https://llmcostlab.com/models/ernie-5-1/"
    },
    {
      "id": "step-3-5-flash",
      "name": "Step 3.5 Flash",
      "provider": "StepFun",
      "inputPerMillionTokens": 0.1,
      "outputPerMillionTokens": 0.3,
      "cachedInputPerMillionTokens": null,
      "batchDiscountPercent": null,
      "blendedPerMillionTokens": 0.15,
      "notes": "Low-cost tier from StepFun, competitive with the cheapest Gemini and Llama options.",
      "providerPricingPage": "https://platform.stepfun.com/",
      "page": "https://llmcostlab.com/models/step-3-5-flash/"
    }
  ]
}