modelmix 5.1.1 → 5.1.4

package/README.md

@@ -117,9 +117,9 @@ const ETH = ModelMix.new()
 console.log(ETH.price);
 ```
 
-**This example uses providers with free quotas (OpenRouter, Groq, Cerebras) - just get the API key and you're ready to go. If one model runs out of quota, ModelMix automatically falls back to the next model in the chain.**
+**This example uses providers with free quotas (Groq, Cerebras, and Together). OpenRouter is disabled because its GPT-OSS 120B route is no longer free. If one model runs out of quota, ModelMix automatically falls back to the next model in the chain.**
 ```javascript
-ModelMix.new()
+ModelMix.new({ mix: { openrouter: false } })
   .gptOss()
   .kimiK25()
   .hermes3()
@@ -175,8 +175,10 @@ ModelMix provides convenient shorthand methods for quickly accessing different A
 | `qwen36plus()` | Fireworks | qwen3p6-plus | [\$0.50][10] | [\$3.00][10] |
 | `qwen37plus()` | Fireworks | models/qwen3p7-plus | [\$0.40][10] | [\$1.60][10] |
 | `qwen38max()` | Fireworks | qwen3p8-2p4t-a95b | [\$2.00][10] | [\$6.00][10] |
+| `qwen3827b()` | OpenRouter | qwen/qwen3.8-27b | [\$0.45][15] | [\$3.20][15] |
 | `deepseekV4Flash()` | Fireworks | models/deepseek-v4-flash | [\$0.14][10] | [\$0.28][10] |
 | `deepseekV4Pro()` | Fireworks | models/deepseek-v4-pro-0813 | [\$1.32][12] | [\$3.96][12] |
+| `GLM53()` | OpenRouter | z-ai/glm-5.3 | [\$1.40][16] | [\$4.40][16] |
 | `GLM52()` | Together | zai-org/GLM-5.2 | [\$1.40][7] | [\$4.40][7] |
 | `GLM51()` | Fireworks | models/glm-5p1 | [\$1.05][10] | [\$3.50][10] |
 | `minimaxM3()` | MiniMax | MiniMax-M3 | [\$0.30][9] | [\$1.20][9] |
@@ -206,6 +208,8 @@ Gemini 3.7 Flash and 3.6 Flash use Google's introductory standard pricing throug
 [12]: https://fireworks.ai/models/deepseek-ai/deepseek-v4-pro-0813 "DeepSeek V4 Pro 0813 Pricing"
 [13]: https://openrouter.ai/nousresearch "Nous Research Models on OpenRouter"
 [14]: https://openrouter.ai/qwen/qwen3.5-397b-a17b "Qwen3.5 397B A17B on OpenRouter"
+[15]: https://openrouter.ai/qwen/qwen3.8-27b "Qwen3.8 27B on OpenRouter"
+[16]: https://openrouter.ai/z-ai/glm-5.3 "GLM 5.3 on OpenRouter"
 
 Each method accepts optional `options`, `config`, and (for multi-provider methods) `mix` parameters to customize behavior.  
 
@@ -246,6 +250,8 @@ ModelMix.new().effort(-1).minimaxM3().addText('...').message();
 
 - **Gemini:** Gemini 3+ uses bands 0–24 / 25–49 / 50–74 / 75–100. Gemini 3.7 Flash clamps these bands to `low` / `low` / `medium` / `high`; `-1` leaves its native `medium` default unchanged. Gemini 2.5 maps 0–100 to `thinkingBudget`.
 - **GPT-5.6:** `100` maps to `max`; 80–99 remains `xhigh`.
+- **Qwen 3.8 27B:** 0–39 / 40–79 / 80–100 map to `low` / `medium` / `xhigh`; `-1` leaves the native `xhigh` default unchanged.
+- **GLM 5.3:** reasoning is mandatory; 0–39 / 40–79 / 80–100 map to `low` / `high` / `max`; `-1` leaves the native `max` default unchanged.
 - **DeepSeek:** `↑` means thinking is enabled; `off` means it is disabled.
 - **MiniMax:** `off` maps to `thinking.disabled`; `adaptive` maps to `thinking.type=adaptive`.
 - **Anthropic:** Claude 5, Fable, Opus 4.6+, and Sonnet 4.6+ use adaptive thinking with `output_config.effort`. Sonnet 4.5 and Haiku 4.5 use `thinking.type=enabled` with `budget_tokens`.

package/demo/free.js

@@ -1,13 +1,12 @@
 import { ModelMix } from '../index.js';
 try { process.loadEnvFile(); } catch {}
 
-const ai = ModelMix.new({ config: { debug: 2 } })
+const ai = ModelMix.new({ config: { debug: 2 }, mix: { openrouter: false } })
     .gptOss()
     .kimiK25()
     .hermes3()
     .addText('What is the capital of France?');
 
 const response = await ai.message();
-console.log('Response from Claude via OpenRouter:', response);
-
+console.log('Response from a free-tier provider:', response);
 

package/demo/short.js

@@ -15,8 +15,10 @@ const mmix = await ModelMix.new(setup)
     .gpt5nano() // (fallback 3) OpenAI gpt-5-nano
     .grok46() // (fallback 4) Grok grok-4.6
     .qwen35397b() // (fallback 5) OpenRouter qwen/qwen3.5-397b-a17b
-    .hermes470b() // (fallback 6) OpenRouter nousresearch/hermes-4-70b
-    .hermes4405b() // (fallback 7) OpenRouter nousresearch/hermes-4-405b
+    .qwen3827b() // (fallback 6) OpenRouter qwen/qwen3.8-27b
+    .GLM53() // (fallback 7) OpenRouter z-ai/glm-5.3
+    .hermes470b() // (fallback 8) OpenRouter nousresearch/hermes-4-70b
+    .hermes4405b() // (fallback 9) OpenRouter nousresearch/hermes-4-405b
     .addText("What's your name?");
 
 console.log(await mmix.message());

package/effort.js

@@ -34,12 +34,20 @@ const GEMINI_BANDS = [
     [75, 100, 'high'],
 ];
 
+const GLM53_BANDS = [
+    [0, 39, 'low'],
+    [40, 79, 'high'],
+    [80, 100, 'max'],
+];
+
 /** Exact model → supported OpenAI reasoning_effort values */
 const OPENAI_MODEL_LEVELS = {
     'gpt-5.6-sol': ['none', 'low', 'medium', 'high', 'xhigh', 'max'],
     'gpt-5.6-terra': ['none', 'low', 'medium', 'high', 'xhigh', 'max'],
     'gpt-5.6-luna': ['none', 'low', 'medium', 'high', 'xhigh', 'max'],
     'accounts/fireworks/models/qwen3p8-2p4t-a95b': ['none', 'low', 'medium', 'high'],
+    'qwen/qwen3.8-27b': ['low', 'medium', 'xhigh'],
+    'z-ai/glm-5.3': ['low', 'high', 'max'],
     'grok-4.6': ['low', 'medium', 'high', 'xhigh'],
     'gpt-5': ['minimal', 'low', 'medium', 'high'],
     'gpt-5-mini': ['minimal', 'low', 'medium', 'high'],
@@ -47,7 +55,6 @@ const OPENAI_MODEL_LEVELS = {
     'gpt-5.3-codex': ['low', 'medium', 'high', 'xhigh'],
     'gpt-oss-120b': ['low', 'medium', 'high'],
     'openai/gpt-oss-120b': ['low', 'medium', 'high'],
-    'openai/gpt-oss-120b:free': ['low', 'medium', 'high'],
 };
 
 /**
@@ -378,9 +385,11 @@ function mapEffort(providerFamily, effort, modelKey) {
             return null;
         }
         const supported = supportedOpenAILevels(modelKey);
-        const desired = normalized === 100 && supported.includes('max')
-            ? 'max'
-            : levelFromBands(normalized, OPENAI_BANDS);
+        const desired = modelKey === 'z-ai/glm-5.3'
+            ? levelFromBands(normalized, GLM53_BANDS)
+            : normalized === 100 && supported.includes('max')
+                ? 'max'
+                : levelFromBands(normalized, OPENAI_BANDS);
         const level = pickNearestLevel(desired, OPENAI_LEVEL_LADDER, supported);
         return { reasoning_effort: level };
     }

package/index.d.ts

@@ -501,6 +501,7 @@ export declare class ModelMix {
   qwen36plus(args?: ModelAttachArgs): this;
   qwen37plus(args?: ModelAttachArgs): this;
   qwen38max(args?: ModelAttachArgs): this;
+  qwen3827b(args?: ModelAttachArgs): this;
   hermes470b(args?: ModelAttachArgs): this;
   hermes4405b(args?: ModelAttachArgs): this;
   hermes3(args?: ModelAttachArgs): this;
@@ -517,6 +518,7 @@ export declare class ModelMix {
   deepseekV4Flash(args?: ModelAttachArgs): this;
   GLM51(args?: ModelAttachArgs): this;
   GLM52(args?: ModelAttachArgs): this;
+  GLM53(args?: ModelAttachArgs): this;
 
   addText(text: string, options?: RoleOptions): this;
   addTextFromFile(filePath: string, options?: RoleOptions): this;