diff --git a/bun.lock b/bun.lock
index 33264cb29..e6d420db5 100644
--- a/bun.lock
+++ b/bun.lock
@@ -381,7 +381,7 @@
     },
     "packages/tfcode": {
       "name": "tfcode",
-      "version": "1.0.7",
+      "version": "1.0.8",
       "bin": {
         "tfcode": "./bin/tfcode",
       },
diff --git a/packages/tfcode/package.json b/packages/tfcode/package.json
index d51733857..ac592e701 100644
--- a/packages/tfcode/package.json
+++ b/packages/tfcode/package.json
@@ -1,6 +1,6 @@
 {
   "$schema": "https://json.schemastore.org/package.json",
-  "version": "1.0.7",
+  "version": "1.0.8",
   "name": "tfcode",
   "type": "module",
   "license": "MIT",
diff --git a/packages/tfcode/src/session/compaction.ts b/packages/tfcode/src/session/compaction.ts
index 072ea1d57..764d2fbab 100644
--- a/packages/tfcode/src/session/compaction.ts
+++ b/packages/tfcode/src/session/compaction.ts
@@ -29,6 +29,7 @@ export namespace SessionCompaction {
   }
 
   const COMPACTION_BUFFER = 20_000
+  const COMPACTION_BUFFER_TOOTHFAIRYAI = 100_000
 
   export async function isOverflow(input: { tokens: MessageV2.Assistant["tokens"]; model: Provider.Model }) {
     const config = await Config.get()
@@ -40,8 +41,11 @@ export namespace SessionCompaction {
       input.tokens.total ||
       input.tokens.input + input.tokens.output + input.tokens.cache.read + input.tokens.cache.write
 
-    const reserved =
-      config.compaction?.reserved ?? Math.min(COMPACTION_BUFFER, ProviderTransform.maxOutputTokens(input.model))
+    const defaultBuffer =
+      input.model.providerID === "toothfairyai"
+        ? COMPACTION_BUFFER_TOOTHFAIRYAI
+        : Math.min(COMPACTION_BUFFER, ProviderTransform.maxOutputTokens(input.model))
+    const reserved = config.compaction?.reserved ?? defaultBuffer
     const usable = input.model.limit.input
       ? input.model.limit.input - reserved
       : context - ProviderTransform.maxOutputTokens(input.model)