floe-guard 0.15.0 → 0.15.2

package/README.md

@@ -64,6 +64,24 @@ const adv = guard.advisory();
 const model = adv.nearLimit ? openai("gpt-4o-mini") : openai("gpt-4o");
 ```
 
+## Request-sized estimates
+
+To ensure the ceiling is enforced on the first run or for a call much larger than the previous one, you can price the actual incoming request using `estimateCall()` and pass the estimate to `reserve()` or `check()`:
+
+```ts
+const est = guard.estimateCall("gpt-4o", 12_000, 4_096);
+const handle = guard.reserve(est); // throws BudgetExceeded NOW if this call alone would cross
+try {
+  const response = await callYourLlm({ model: "gpt-4o", ... });
+  guard.settle("gpt-4o", response.usage.promptTokens, response.usage.completionTokens, { reserved: handle });
+} catch (err) {
+  guard.release(handle);
+  throw err;
+}
+```
+
+If the model is unpriceable, `estimateCall()` returns `undefined` and `reserve(undefined)` / `check(undefined)` fall back gracefully to the last-cost prediction.
+
 ## Tool spend under the same ceiling
 
 Paid tool calls (Apollo, Exa, scrapers) burn the same budget as tokens. The

package/dist/adapters/livekit.cjs

@@ -76,15 +76,40 @@ function preCall(guard, options = {}) {
 // src/cost_map.json
 var cost_map_default = {
   __meta__: {
-    generated_at: "2026-08-17",
+    generated_at: "2026-08-24",
     source: "LiteLLM public model prices (bundled snapshot); voice rates verified from vendor list pages"
   },
+  "chat-latest": {
+    input_cost_per_token: 5e-6,
+    output_cost_per_token: 3e-5,
+    litellm_provider: "openai",
+    mode: "chat",
+    cache_read_input_token_cost: 5e-7
+  },
   "chatgpt-4o-latest": {
     input_cost_per_token: 5e-6,
     output_cost_per_token: 15e-6,
     litellm_provider: "openai",
     mode: "chat"
   },
+  "claude-3-5-haiku-20241022": {
+    input_cost_per_token: 8e-7,
+    output_cost_per_token: 4e-6,
+    litellm_provider: "anthropic",
+    mode: "chat"
+  },
+  "claude-3-5-sonnet-20240620": {
+    input_cost_per_token: 3e-6,
+    output_cost_per_token: 15e-6,
+    litellm_provider: "anthropic",
+    mode: "chat"
+  },
+  "claude-3-5-sonnet-20241022": {
+    input_cost_per_token: 3e-6,
+    output_cost_per_token: 15e-6,
+    litellm_provider: "anthropic",
+    mode: "chat"
+  },
   "claude-3-7-sonnet-20250219": {
     input_cost_per_token: 3e-6,
     output_cost_per_token: 15e-6,
@@ -293,6 +318,22 @@ var cost_map_default = {
     cache_read_input_token_cost: 2e-7,
     cache_creation_input_token_cost: 25e-7
   },
+  "daybreak-blue-latest": {
+    input_cost_per_token: 4e-6,
+    output_cost_per_token: 2e-5,
+    litellm_provider: "openai",
+    mode: "chat",
+    cache_read_input_token_cost: 4e-7,
+    cache_creation_input_token_cost: 5e-6
+  },
+  "daybreak-red-latest": {
+    input_cost_per_token: 125e-7,
+    output_cost_per_token: 75e-6,
+    litellm_provider: "openai",
+    mode: "chat",
+    cache_read_input_token_cost: 125e-8,
+    cache_creation_input_token_cost: 15625e-9
+  },
   "ft:gpt-3.5-turbo": {
     input_cost_per_token: 3e-6,
     output_cost_per_token: 6e-6,
@@ -384,7 +425,7 @@ var cost_map_default = {
     output_cost_per_token: 4e-7,
     litellm_provider: "gemini",
     mode: "chat",
-    cache_read_input_token_cost: 375e-10
+    cache_read_input_token_cost: 25e-9
   },
   "gemini-2.0-flash-lite": {
     input_cost_per_token: 75e-9,
@@ -535,12 +576,18 @@ var cost_map_default = {
     mode: "chat",
     cache_read_input_token_cost: 3e-8
   },
+  "gemini-3.5-live-translate-preview": {
+    input_cost_per_token: 35e-7,
+    output_cost_per_token: 21e-6,
+    litellm_provider: "gemini",
+    mode: "chat"
+  },
   "gemini-3.6-flash": {
-    input_cost_per_token: 15e-7,
-    output_cost_per_token: 75e-7,
+    input_cost_per_token: 75e-8,
+    output_cost_per_token: 375e-8,
     litellm_provider: "gemini",
     mode: "chat",
-    cache_read_input_token_cost: 15e-8
+    cache_read_input_token_cost: 75e-9
   },
   "gemini-3.7-flash": {
     input_cost_per_token: 75e-8,
@@ -1027,12 +1074,20 @@ var cost_map_default = {
     cache_read_input_token_cost: 5e-7
   },
   "gpt-5.6": {
-    input_cost_per_token: 5e-6,
-    output_cost_per_token: 3e-5,
+    input_cost_per_token: 4e-6,
+    output_cost_per_token: 2e-5,
     litellm_provider: "openai",
     mode: "chat",
-    cache_read_input_token_cost: 5e-7,
-    cache_creation_input_token_cost: 625e-8
+    cache_read_input_token_cost: 4e-7,
+    cache_creation_input_token_cost: 5e-6
+  },
+  "gpt-5.6-cyber": {
+    input_cost_per_token: 125e-7,
+    output_cost_per_token: 75e-6,
+    litellm_provider: "openai",
+    mode: "chat",
+    cache_read_input_token_cost: 125e-8,
+    cache_creation_input_token_cost: 15625e-9
   },
   "gpt-5.6-luna": {
     input_cost_per_token: 2e-7,
@@ -1043,12 +1098,12 @@ var cost_map_default = {
     cache_creation_input_token_cost: 25e-8
   },
   "gpt-5.6-sol": {
-    input_cost_per_token: 5e-6,
-    output_cost_per_token: 3e-5,
+    input_cost_per_token: 4e-6,
+    output_cost_per_token: 2e-5,
     litellm_provider: "openai",
     mode: "chat",
-    cache_read_input_token_cost: 5e-7,
-    cache_creation_input_token_cost: 625e-8
+    cache_read_input_token_cost: 4e-7,
+    cache_creation_input_token_cost: 5e-6
   },
   "gpt-5.6-terra": {
     input_cost_per_token: 2e-6,
@@ -1298,24 +1353,6 @@ var cost_map_default = {
       rate: 4e-3,
       provider: "twilio"
     }
-  },
-  "claude-3-5-sonnet-20241022": {
-    input_cost_per_token: 3e-6,
-    output_cost_per_token: 15e-6,
-    litellm_provider: "anthropic",
-    mode: "chat"
-  },
-  "claude-3-5-sonnet-20240620": {
-    input_cost_per_token: 3e-6,
-    output_cost_per_token: 15e-6,
-    litellm_provider: "anthropic",
-    mode: "chat"
-  },
-  "claude-3-5-haiku-20241022": {
-    input_cost_per_token: 8e-7,
-    output_cost_per_token: 4e-6,
-    litellm_provider: "anthropic",
-    mode: "chat"
   }
 };