Implement switch to glm-5.1 for free mode

Author: jahooma

agents/__tests__/editor.test.ts

@@ -62,9 +62,9 @@ describe('editor agent', () => {
       expect(gpt5Editor.model).toBe('openai/gpt-5.1')
     })
 
-    test('creates minimax editor', () => {
-      const minimaxEditor = createCodeEditor({ model: 'minimax' })
-      expect(minimaxEditor.model).toBe('minimax/minimax-m2.5')
+    test('creates glm editor', () => {
+      const glmEditor = createCodeEditor({ model: 'glm' })
+      expect(glmEditor.model).toBe('z-ai/glm-5.1')
     })
 
     test('gpt-5 editor does not include think tags in instructions', () => {
@@ -74,9 +74,9 @@ describe('editor agent', () => {
     })
 
     test('glm editor does not include think tags in instructions', () => {
-      const minimaxEditor = createCodeEditor({ model: 'minimax' })
-      expect(minimaxEditor.instructionsPrompt).not.toContain('<think>')
-      expect(minimaxEditor.instructionsPrompt).not.toContain('</think>')
+      const glmEditor = createCodeEditor({ model: 'glm' })
+      expect(glmEditor.instructionsPrompt).not.toContain('<think>')
+      expect(glmEditor.instructionsPrompt).not.toContain('</think>')
     })
 
     test('opus editor includes think tags in instructions', () => {
@@ -88,17 +88,17 @@ describe('editor agent', () => {
     test('all variants have same base properties', () => {
       const opusEditor = createCodeEditor({ model: 'opus' })
       const gpt5Editor = createCodeEditor({ model: 'gpt-5' })
-      const minimaxEditor = createCodeEditor({ model: 'minimax' })
+      const glmEditor = createCodeEditor({ model: 'glm' })
 
       // All should have same basic structure
       expect(opusEditor.displayName).toBe(gpt5Editor.displayName)
-      expect(gpt5Editor.displayName).toBe(minimaxEditor.displayName)
+      expect(gpt5Editor.displayName).toBe(glmEditor.displayName)
 
       expect(opusEditor.outputMode).toBe(gpt5Editor.outputMode)
-      expect(gpt5Editor.outputMode).toBe(minimaxEditor.outputMode)
+      expect(gpt5Editor.outputMode).toBe(glmEditor.outputMode)
 
       expect(opusEditor.toolNames).toEqual(gpt5Editor.toolNames)
-      expect(gpt5Editor.toolNames).toEqual(minimaxEditor.toolNames)
+      expect(gpt5Editor.toolNames).toEqual(glmEditor.toolNames)
     })
   })
 

agents/base2/base2.ts

@@ -28,7 +28,7 @@ export function createBase2(
 
   return {
     publisher,
-    model: isFree ? 'minimax/minimax-m2.5' : 'anthropic/claude-opus-4.6',
+    model: isFree ? 'z-ai/glm-5.1' : 'anthropic/claude-opus-4.6',
     providerOptions: isFree ? {
       data_collection: 'deny',
     } : {

agents/editor/editor-lite.ts

@@ -3,7 +3,7 @@ import { createCodeEditor } from './editor'
 import type { AgentDefinition } from '../types/agent-definition'
 
 const definition: AgentDefinition = {
-  ...createCodeEditor({ model: 'minimax' }),
+  ...createCodeEditor({ model: 'glm' }),
   id: 'editor-lite',
 }
 export default definition

agents/editor/editor.ts

@@ -4,16 +4,16 @@ import { publisher } from '../constants'
 import type { AgentDefinition } from '../types/agent-definition'
 
 export const createCodeEditor = (options: {
-  model: 'gpt-5' | 'opus' | 'minimax'
+  model: 'gpt-5' | 'opus' | 'glm'
 }): Omit<AgentDefinition, 'id'> => {
   const { model } = options
   return {
     publisher,
     model:
       options.model === 'gpt-5'
         ? 'openai/gpt-5.1'
-        : options.model === 'minimax'
-          ? 'minimax/minimax-m2.5'
+        : options.model === 'glm'
+          ? 'z-ai/glm-5.1'
           : 'anthropic/claude-opus-4.6',
     ...(options.model === 'opus' && {
       providerOptions: {
@@ -65,7 +65,7 @@ OR for new files or major rewrites:
 }
 </codebuff_tool_call>
 
-${model === 'gpt-5' || model === 'minimax'
+${model === 'gpt-5' || model === 'glm'
         ? ''
         : `Before you start writing your implementation, you should use <think> tags to think about the best way to implement the changes.
 

agents/reviewer/code-reviewer-lite.ts

@@ -5,7 +5,7 @@ import { createReviewer } from './code-reviewer'
 const definition: SecretAgentDefinition = {
   id: 'code-reviewer-lite',
   publisher,
-  ...createReviewer('minimax/minimax-m2.5'),
+  ...createReviewer('z-ai/glm-5.1'),
 }
 
 export default definition

agents/types/agent-definition.ts

@@ -424,6 +424,7 @@ export type ModelName =
   | 'moonshotai/kimi-k2.5'
   | 'moonshotai/kimi-k2.5:nitro'
   | 'z-ai/glm-5'
+  | 'z-ai/glm-5.1'
   | 'z-ai/glm-4.6'
   | 'z-ai/glm-4.6:nitro'
   | 'z-ai/glm-4.7'

common/src/constants/free-agents.ts

@@ -18,7 +18,7 @@ export const FREE_COST_MODE = 'free' as const
  */
 export const FREE_MODE_AGENT_MODELS: Record<string, Set<string>> = {
   // Root orchestrator
-  'base2-free': new Set(['minimax/minimax-m2.5']),
+  'base2-free': new Set(['minimax/minimax-m2.5', 'z-ai/glm-5.1']),
 
   // File exploration agents
   'file-picker': new Set(['google/gemini-2.5-flash-lite']),
@@ -33,10 +33,10 @@ export const FREE_MODE_AGENT_MODELS: Record<string, Set<string>> = {
   'basher': new Set(['google/gemini-3.1-flash-lite-preview']),
 
   // Editor for free mode
-  'editor-lite': new Set(['minimax/minimax-m2.5']),
+  'editor-lite': new Set(['minimax/minimax-m2.5', 'z-ai/glm-5.1']),
 
   // Code reviewer for free mode
-  'code-reviewer-lite': new Set(['minimax/minimax-m2.5']),
+  'code-reviewer-lite': new Set(['minimax/minimax-m2.5', 'z-ai/glm-5.1']),
 
   // Thinker for free mode
   'thinker-with-files-gemini': new Set(['google/gemini-3.1-pro-preview']),

common/src/templates/initial-agents-dir/types/agent-definition.ts

@@ -424,6 +424,7 @@ export type ModelName =
   | 'moonshotai/kimi-k2.5'
   | 'moonshotai/kimi-k2.5:nitro'
   | 'z-ai/glm-5'
+  | 'z-ai/glm-5.1'
   | 'z-ai/glm-4.6'
   | 'z-ai/glm-4.6:nitro'
   | 'z-ai/glm-4.7'

web/src/llm-api/fireworks.ts

@@ -29,6 +29,7 @@ const fireworksAgent = new Agent({
 /** Map from OpenRouter model IDs to Fireworks standard API model IDs */
 const FIREWORKS_MODEL_MAP: Record<string, string> = {
   'minimax/minimax-m2.5': 'accounts/fireworks/models/minimax-m2p5',
+  'z-ai/glm-5.1': 'accounts/fireworks/models/glm-5p1',
 }
 
 /** Flag to enable custom Fireworks deployments (set to false to use global API only) */
@@ -137,12 +138,31 @@ function createFireworksRequest(params: {
   })
 }
 
-// Fireworks per-token pricing (dollars per token)
-const FIREWORKS_INPUT_COST_PER_TOKEN = 0.30 / 1_000_000
-const FIREWORKS_CACHED_INPUT_COST_PER_TOKEN = 0.03 / 1_000_000
-const FIREWORKS_OUTPUT_COST_PER_TOKEN = 1.20 / 1_000_000
+// Fireworks per-token pricing (dollars per token), keyed by OpenRouter model ID
+interface FireworksPricing {
+  inputCostPerToken: number
+  cachedInputCostPerToken: number
+  outputCostPerToken: number
+}
+
+const FIREWORKS_PRICING_MAP: Record<string, FireworksPricing> = {
+  'minimax/minimax-m2.5': {
+    inputCostPerToken: 0.30 / 1_000_000,
+    cachedInputCostPerToken: 0.03 / 1_000_000,
+    outputCostPerToken: 1.20 / 1_000_000,
+  },
+  'z-ai/glm-5.1': {
+    inputCostPerToken: 1.40 / 1_000_000,
+    cachedInputCostPerToken: 0.26 / 1_000_000,
+    outputCostPerToken: 4.40 / 1_000_000,
+  },
+}
+
+function getFireworksPricing(model: string): FireworksPricing {
+  return FIREWORKS_PRICING_MAP[model] ?? FIREWORKS_MODEL_MAP['z-ai/glm-5.1']
+}
 
-function extractUsageAndCost(usage: Record<string, unknown> | undefined | null): UsageData {
+function extractUsageAndCost(usage: Record<string, unknown> | undefined | null, model: string): UsageData {
   if (!usage) return { inputTokens: 0, outputTokens: 0, cacheReadInputTokens: 0, reasoningTokens: 0, cost: 0 }
   const promptDetails = usage.prompt_tokens_details as Record<string, unknown> | undefined | null
   const completionDetails = usage.completion_tokens_details as Record<string, unknown> | undefined | null
@@ -153,11 +173,12 @@ function extractUsageAndCost(usage: Record<string, unknown> | undefined | null):
   const reasoningTokens = typeof completionDetails?.reasoning_tokens === 'number' ? completionDetails.reasoning_tokens : 0
 
   // Fireworks doesn't return cost — compute from token counts and known pricing
+  const pricing = getFireworksPricing(model)
   const nonCachedInputTokens = Math.max(0, inputTokens - cacheReadInputTokens)
   const cost =
-    nonCachedInputTokens * FIREWORKS_INPUT_COST_PER_TOKEN +
-    cacheReadInputTokens * FIREWORKS_CACHED_INPUT_COST_PER_TOKEN +
-    outputTokens * FIREWORKS_OUTPUT_COST_PER_TOKEN
+    nonCachedInputTokens * pricing.inputCostPerToken +
+    cacheReadInputTokens * pricing.cachedInputCostPerToken +
+    outputTokens * pricing.outputCostPerToken
 
   return { inputTokens, outputTokens, cacheReadInputTokens, reasoningTokens, cost }
 }
@@ -192,7 +213,7 @@ export async function handleFireworksNonStream({
   const data = await response.json()
   const content = data.choices?.[0]?.message?.content ?? ''
   const reasoningText = data.choices?.[0]?.message?.reasoning_content ?? data.choices?.[0]?.message?.reasoning ?? ''
-  const usageData = extractUsageAndCost(data.usage)
+  const usageData = extractUsageAndCost(data.usage, originalModel)
 
   insertMessageToBigQuery({
     messageId: data.id,
@@ -493,7 +514,7 @@ async function handleResponse({
     return { state }
   }
 
-  const usageData = extractUsageAndCost(data.usage as Record<string, unknown>)
+  const usageData = extractUsageAndCost(data.usage as Record<string, unknown>, originalModel)
   const messageId = typeof data.id === 'string' ? data.id : 'unknown'
 
   insertMessageToBigQuery({