{
  "$schema": "./pricing.schema.json",
  "source": "https://agentshortlist.com",
  "sourceRepo": "https://github.com/lucaspowell8020/agentshortlist",
  "verifiedDate": "2026-04-27",
  "verificationCadence": "daily",
  "verificationMethod": "Pricing is verified daily against vendor sources. Small drift (under 25% on input and output) is auto-applied; larger changes are reviewed by a human editor before publication. Vendor pricing pages are the canonical source of truth.",
  "license": "CC-BY-4.0",
  "models": [
    {
      "slug": "claude-opus-4-7",
      "name": "Claude Opus 4.7",
      "vendor": "Anthropic",
      "inputPricePerMillion": 5,
      "outputPricePerMillion": 25,
      "contextWindow": 1000000,
      "tier": "frontier",
      "notes": "Best-in-class reasoning. Anthropic's flagship for the hardest agentic decisions."
    },
    {
      "slug": "claude-sonnet-4-6",
      "name": "Claude Sonnet 4.6",
      "vendor": "Anthropic",
      "inputPricePerMillion": 3,
      "outputPricePerMillion": 15,
      "contextWindow": 1000000,
      "tier": "balanced",
      "notes": "The sweet spot. Handles ~90% of agent workflows at a fraction of Opus pricing."
    },
    {
      "slug": "claude-haiku-4-5",
      "name": "Claude Haiku 4.5",
      "vendor": "Anthropic",
      "inputPricePerMillion": 1,
      "outputPricePerMillion": 5,
      "contextWindow": 200000,
      "tier": "value",
      "notes": "Fast and cheap. Best for high-volume classification and short replies."
    },
    {
      "slug": "gpt-5",
      "name": "GPT-5.5",
      "vendor": "OpenAI",
      "inputPricePerMillion": 5,
      "outputPricePerMillion": 30,
      "contextWindow": 400000,
      "tier": "frontier",
      "notes": "OpenAI's frontier line. Marketed for coding and professional work."
    },
    {
      "slug": "gpt-5-mini",
      "name": "GPT-5.4 mini",
      "vendor": "OpenAI",
      "inputPricePerMillion": 0.75,
      "outputPricePerMillion": 4.5,
      "contextWindow": 128000,
      "tier": "value"
    },
    {
      "slug": "grok-4-20",
      "name": "Grok 4.20",
      "vendor": "xAI",
      "inputPricePerMillion": 2,
      "outputPricePerMillion": 6,
      "contextWindow": 2000000,
      "tier": "frontier",
      "notes": "xAI's current flagship. 2M context with live web access via X integration."
    },
    {
      "slug": "grok-4",
      "name": "Grok 4",
      "vendor": "xAI",
      "inputPricePerMillion": 3,
      "outputPricePerMillion": 15,
      "contextWindow": 256000,
      "tier": "frontier",
      "notes": "xAI's previous flagship. Still available; most builders should default to Grok 4.20."
    },
    {
      "slug": "grok-4-fast",
      "name": "Grok 4 Fast",
      "vendor": "xAI",
      "inputPricePerMillion": 0.2,
      "outputPricePerMillion": 0.5,
      "contextWindow": 2000000,
      "tier": "value",
      "notes": "xAI's value tier. 10× cheaper than Grok 4.20 with the same 2M context."
    },
    {
      "slug": "gemini-2-5-pro",
      "name": "Gemini 2.5 Pro",
      "vendor": "Google",
      "inputPricePerMillion": 1.25,
      "outputPricePerMillion": 10,
      "contextWindow": 2000000,
      "tier": "balanced",
      "notes": "Largest context window on the list. Pricing shown is for prompts ≤200k tokens; rises to $2.50/$15 above that."
    },
    {
      "slug": "gemini-2-5-flash",
      "name": "Gemini 2.5 Flash",
      "vendor": "Google",
      "inputPricePerMillion": 0.3,
      "outputPricePerMillion": 2.5,
      "contextWindow": 1000000,
      "tier": "value",
      "notes": "Hybrid reasoning model with 1M context and tunable thinking budgets."
    },
    {
      "slug": "gemini-3-1-pro-preview",
      "name": "Gemini 3.1 Pro Preview",
      "vendor": "Google",
      "inputPricePerMillion": 2,
      "outputPricePerMillion": 12,
      "contextWindow": 2000000,
      "tier": "balanced"
    },
    {
      "slug": "gemini-3-1-flash-lite-preview",
      "name": "Gemini 3.1 Flash-Lite Preview",
      "vendor": "Google",
      "inputPricePerMillion": 0.25,
      "outputPricePerMillion": 1.5,
      "contextWindow": 1000000,
      "tier": "value",
      "notes": "Google's most cost-efficient model. Optimised for high-volume agentic tasks and simple data processing."
    },
    {
      "slug": "llama-3-3-70b",
      "name": "Llama 3.3 70B (Together AI)",
      "vendor": "Meta / Together AI",
      "inputPricePerMillion": 0.88,
      "outputPricePerMillion": 0.88,
      "contextWindow": 128000,
      "tier": "value",
      "notes": "Open weights. The price shown reflects Together AI; other inference providers host the same weights at different rates."
    },
    {
      "slug": "deepseek-v4-flash",
      "name": "DeepSeek V4 Flash",
      "vendor": "DeepSeek",
      "inputPricePerMillion": 0.14,
      "outputPricePerMillion": 0.28,
      "contextWindow": 128000,
      "tier": "value",
      "notes": "DeepSeek's value tier. Cache-hit pricing drops input to $0.0028/M on repeat queries."
    },
    {
      "slug": "kimi-k2",
      "name": "Kimi K2",
      "vendor": "Moonshot AI",
      "inputPricePerMillion": 0.57,
      "outputPricePerMillion": 2.3,
      "contextWindow": 200000,
      "tier": "value",
      "notes": "Moonshot AI's frontier model. Strong on long-context work."
    },
    {
      "slug": "glm-4-6",
      "name": "GLM-4.6",
      "vendor": "z.ai",
      "inputPricePerMillion": 0.6,
      "outputPricePerMillion": 2.2,
      "contextWindow": 128000,
      "tier": "value",
      "notes": "Formerly Zhipu AI. Competitive Chinese frontier model with strong coding and tool-use benchmarks."
    },
    {
      "slug": "mistral-large",
      "name": "Mistral Large 2.1",
      "vendor": "Mistral",
      "inputPricePerMillion": 2,
      "outputPricePerMillion": 6,
      "contextWindow": 128000,
      "tier": "balanced",
      "notes": "European frontier model. Stronger data residency story than US providers — useful for EU compliance-heavy workflows."
    }
  ]
}