diff --git a/lib/estimate.js b/lib/estimate.js
index 14de8b9..3559d40 100644
--- a/lib/estimate.js
+++ b/lib/estimate.js
@@ -90,6 +90,40 @@ const SPEECH_RATE = {
 const SPEECH_DEFAULT_RATE = 0.05; // elevenlabs/flash-v2.5 (the MCP default)
 const SOUND_EFFECT_COST = 0.0525;
 
+// Chat models billed $0 — mirrors @blockrun/mcp FREE_CHAT_MODELS
+// (src/utils/constants.ts), refreshed 2026-10-20. Membership by id, not
+// vendor: cohere/ and poolside/ have $0 models, and not every nvidia/ id is
+// guaranteed free. Sources: GET /v1/models billing_mode "free" and the account
+// price sheet (blockrun.ai/api/pricing) at $0.
+const FREE_CHAT_MODELS = new Set([
+  "nvidia/gpt-oss-20b", "nvidia/nemotron-3-super-120b", "nvidia/llama-3.2-11b-vision",
+  "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning", "cohere/north-mini-code", "poolside/laguna-xs-2.1",
+  "nvidia/gpt-oss-120b", "nvidia/nemotron-3-ultra-550b", "nvidia/step-3.7-flash", "nvidia/mistral-nemotron",
+  "nvidia/nemotron-nano-12b-v2-vl", "nvidia/nemotron-nano-9b-v2", "nvidia/nemotron-3.5-lightning",
+  "nvidia/muse-glimmer-30b", "nvidia/gemma-4-31b", "nvidia/nemotron-3-super-120b-a12b-free",
+  "nvidia/qwen3.5-397b-a17b", "nvidia/nemotron-ultra-253b", "nvidia/nemotron-super-49b", "nvidia/deepseek-v3.2",
+  "nvidia/nemotron-3-nano-omni-nim",
+  "nvidia/mistral-large-3-675b", "nvidia/qwen3-coder-480b", "nvidia/devstral-2-123b", "nvidia/glm-4.7",
+  "nvidia/llama-4-maverick", "nvidia/qwen3-next-80b-a3b-instruct", "nvidia/qwen3.5-122b-a10b",
+  "nvidia/seed-oss-36b", "nvidia/mistral-small-4-119b",
+]);
+// nvidia/deepseek-v4-pro and nvidia/deepseek-v4-flash are NOT here: the
+// gateways retired both (unknown-model 400; blockrun#848 and its follow-up).
+//
+// Bare spellings the gateway also accepts. A free id whose bare name belongs
+// to a PAID model must be listed here and excluded, so the paid model wins
+// the spelling (MCP FREE_SHADOWED_BY_PAID). Empty since the nvidia DeepSeek
+// mirror was retired; bare "deepseek-v4-pro" is the paid
+// deepseek/deepseek-v4-pro ($1.32/$3.96) and is not in the free set at all.
+const FREE_SHADOWED_BY_PAID = new Set([]);
+const FREE_CHAT_BARE = new Set(
+  [...FREE_CHAT_MODELS].filter((id) => !FREE_SHADOWED_BY_PAID.has(id)).map((id) => id.slice(id.indexOf("/") + 1)),
+);
+function isFreeChatModel(model) {
+  const id = String(model || "").trim();
+  return FREE_CHAT_MODELS.has(id) || (!id.includes("/") && FREE_CHAT_BARE.has(id));
+}
+
 // Whether `size` is in `model`'s large tier (see IMAGE_LARGE_THRESHOLD /
 // IMAGE_LARGE_MIN_DIM). Unparseable sizes price at the base.
 function isLargerThanBase(model, size) {
@@ -147,7 +181,9 @@ function estimate(toolName, input = {}) {
       return { usd: paid ? 0.001 : 0, label: `price · ${cat || "crypto"}${paid ? "" : " (free)"}` };
     }
     case "chat": {
-      const free = input.mode === "free" || /^nvidia\//.test(input.model || "");
+      // An explicit model wins over mode, as in the MCP: mode is ignored when
+      // model is set.
+      const free = input.model ? isFreeChatModel(input.model) : input.mode === "free";
       return { usd: free ? 0 : 0.005, label: `chat${free ? " (free tier)" : ""}` };
     }
     // Wallet + models are always free; data tools (if a fuller profile is used)