Skip to content

Voice-to-text dictation app with local (Nvidia Parakeet/Whisper) and cloud models (BYOK). Privacy-first and available cross-platform.

feat: add astra and fable 5.1 to reasoning models (#2107)

Chad Piha1c65e90e7a7a
diff --git a/CLAUDE.md b/CLAUDE.md
index b87f7b98c..9daf89b25 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -355,10 +355,12 @@ Non-secret env vars persisted to `.env` (via `saveAllKeysToEnvFile()`):
### 6. Language Support
-58 languages supported (see src/utils/languages.ts):
+60 languages supported, defined in `src/config/languageRegistry.json` and read through
+`src/utils/languageSupport.ts`:
-- Each language has a two-letter code and label
-- "auto" for automatic detection
+- Each entry has a two-letter code, label, flag, and per-engine support flags
+ (`whisper`, `parakeet`, `assemblyai`) — 59 Whisper, 26 Parakeet, 7 AssemblyAI
+- "auto" is a 61st entry for automatic detection
- Passed to whisper.cpp via -l parameter
### 7. Agent Naming System
@@ -369,18 +371,27 @@ Non-secret env vars persisted to `.env` (via `saveAllKeysToEnvFile()`):
- Standalone wake-word commands stream into the assistant panel (the address is stripped first, `stripAgentAddress`); a highlighted selection is edited in place by the dictation agent
- Supports multiple AI providers (all models defined in `src/models/modelRegistryData.json`):
- **OpenAI** (Responses API):
- - GPT-5.5 (`gpt-5.5`) - Latest flagship frontier model, 1M context
+ - GPT-6 Astra (`gpt-6-astra`) - Most capable OpenAI model for coding, computer use, and research, 1M context
+ - GPT-5.6 Series (`gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`) - Flagship, balanced, and fastest tiers
+ - GPT-5.5 (`gpt-5.5`) - Frontier model for complex reasoning, 1M context
- GPT-5.2 (`gpt-5.2`) - Strong reasoning model
- GPT-5 Mini (`gpt-5-mini`) - Fast and cost-efficient
- GPT-5 Nano (`gpt-5-nano`) - Ultra-fast, low latency
- GPT-4.1 Series (`gpt-4.1`, `gpt-4.1-mini`, `gpt-4.1-nano`) - Strong baseline with 1M context
+ - GPT-5 and newer reject `temperature`, so their registry entries carry `supportsTemperature: false`
- **Anthropic** (Via IPC bridge to avoid CORS):
- - Claude Opus 4.7 (`claude-opus-4-7`) - Most capable Claude model, 1M context
+ - Claude Fable 5.1 (`claude-fable-5-1`) - Most capable Claude model, Mythos-class, 1M context
+ - Claude Fable 5 (`claude-fable-5`) - Previous Mythos-class flagship, 1M context
+ - Claude Sonnet 5 (`claude-sonnet-5`) - Fast, capable agentic model at lower cost
- Claude Sonnet 4.6 (`claude-sonnet-4-6`) - Balanced performance
- Claude Haiku 4.5 (`claude-haiku-4-5`) - Fast with near-frontier intelligence
+ - Claude Opus 5 (`claude-opus-5`) - Most capable Opus model, 1M context
+ - Claude Opus 4.8 (`claude-opus-4-8`) - Powerful Opus model tuned for honesty and reliability, 1M context
+ - Claude Opus 4.7 (`claude-opus-4-7`) - Powerful Opus model, 1M context
- Claude Opus 4.6 (`claude-opus-4-6`) - Previous Opus generation, 1M context
- Claude Sonnet 4.5 (`claude-sonnet-4-5`) - Previous Sonnet generation
- Claude Opus 4.5 (`claude-opus-4-5`) - Earlier Opus model
+ - Models from Opus 4.7 onward reject `temperature` (`supportsTemperature: false`); the Anthropic IPC bridge sends no thinking parameters, so no Anthropic entry carries `supportsThinking`
- **Google Gemini** (Direct API integration):
- Gemini 3.5 Flash (`gemini-3.5-flash`) - Latest fast, high-capability Gemini model
- Gemini 3.5 Flash Lite (`gemini-3.5-flash-lite`) - Fastest, most cost-effective 3.5 model
@@ -405,9 +416,11 @@ All AI model definitions are centralized in `src/models/modelRegistryData.json`
**Key files:**
- `src/models/modelRegistryData.json` - Single source of truth for all models
-- `src/models/ModelRegistry.ts` - TypeScript wrapper with helper methods
-- `src/config/aiProvidersConfig.ts` - Derives AI_MODES from registry
-- `src/utils/languages.ts` - Derives REASONING_PROVIDERS from registry
+- `src/models/ModelRegistry.ts` - TypeScript wrapper with helper methods; also derives
+ `REASONING_PROVIDERS` (`buildReasoningProviders()`), consumed by the model pickers
+- `src/models/providerDefaultModel.ts` - `pickProviderDefaultModel()`; with no
+ `defaultModel` on the provider, the first model in the array is the default
+- `src/config/retiredCloudModels.ts` - Remaps selections pinned to a retired model
- `src/helpers/modelManagerBridge.js` - Handles local model downloads
**Local model features:**
diff --git a/src/locales/de/translation.json b/src/locales/de/translation.json
index 159cda694..0bd13a8df 100644
--- a/src/locales/de/translation.json
+++ b/src/locales/de/translation.json
@@ -2644,7 +2644,8 @@
"openai_gpt_4_1": "Starke Basis, 1M Kontext",
"openai_gpt_4_1_mini": "Kleineres GPT-4.1-Modell",
"openai_gpt_4_1_nano": "GPT-4.1 mit der niedrigsten Latenz",
- "anthropic_claude_fable_5": "Leistungsstärkstes Modell von Anthropic, Mythos-Klasse",
+ "anthropic_claude_fable_5_1": "Leistungsstärkstes Modell von Anthropic, Mythos-Klasse",
+ "anthropic_claude_fable_5": "Vorheriges Flaggschiff der Mythos-Klasse",
"anthropic_claude_sonnet_5": "Schnelles, leistungsfähiges Assistentenmodell zu geringeren Kosten",
"anthropic_claude_sonnet_4_6": "Ausgewogene Leistung",
"anthropic_claude_sonnet_4_5": "Vorherige Sonnet-Generation",
@@ -2661,6 +2662,7 @@
"groq_openai_gpt_oss_20b": "Schnelles Open-Source-Modell, 1000 T/s",
"groq_compound": "Groqs Compound-System, 450 T/s",
"groq_compound_mini": "Schnelles Compound-System, 3x niedrigere Latenz",
+ "openai_gpt_6_astra": "Leistungsstärkstes Modell von OpenAI für Coding und Forschung",
"openai_gpt_5_6_sol": "Flaggschiff-GPT-5.6 für komplexes Coding und Forschung",
"openai_gpt_5_6_terra": "Ausgewogenes GPT-5.6 für die tägliche Arbeit",
"openai_gpt_5_6_luna": "Schnellstes und günstigstes GPT-5.6",
diff --git a/src/locales/en/translation.json b/src/locales/en/translation.json
index 0adc52dbb..86a585ae8 100644
--- a/src/locales/en/translation.json
+++ b/src/locales/en/translation.json
@@ -2644,7 +2644,8 @@
"openai_gpt_4_1": "Strong baseline, 1M context",
"openai_gpt_4_1_mini": "Smaller GPT-4.1 model",
"openai_gpt_4_1_nano": "Lowest latency GPT-4.1",
- "anthropic_claude_fable_5": "Anthropic's most capable model, Mythos-class",
+ "anthropic_claude_fable_5_1": "Anthropic's most capable model, Mythos-class",
+ "anthropic_claude_fable_5": "Previous Mythos-class flagship",
"anthropic_claude_sonnet_5": "Fast, capable assistant model at lower cost",
"anthropic_claude_sonnet_4_6": "Balanced performance",
"anthropic_claude_sonnet_4_5": "Previous Sonnet generation",
@@ -2661,6 +2662,7 @@
"groq_openai_gpt_oss_20b": "Fast open-source model, 1000 T/sec",
"groq_compound": "Groq's compound system, 450 T/sec",
"groq_compound_mini": "Fast compound system, 3x lower latency",
+ "openai_gpt_6_astra": "OpenAI's most capable model for coding and research",
"openai_gpt_5_6_sol": "Flagship GPT-5.6 for complex coding and research",
"openai_gpt_5_6_terra": "Balanced GPT-5.6 for everyday work",
"openai_gpt_5_6_luna": "Fastest, lowest-cost GPT-5.6",
diff --git a/src/locales/es/translation.json b/src/locales/es/translation.json
index 8c83bc00c..b6ca7be33 100644
--- a/src/locales/es/translation.json
+++ b/src/locales/es/translation.json
@@ -2599,7 +2599,8 @@
"openai_gpt_4_1": "Base sólida, contexto de 1M",
"openai_gpt_4_1_mini": "Modelo GPT-4.1 más pequeño",
"openai_gpt_4_1_nano": "GPT-4.1 de menor latencia",
- "anthropic_claude_fable_5": "El modelo más capaz de Anthropic, clase Mythos",
+ "anthropic_claude_fable_5_1": "El modelo más capaz de Anthropic, clase Mythos",
+ "anthropic_claude_fable_5": "Modelo insignia anterior de clase Mythos",
"anthropic_claude_sonnet_5": "Modelo asistente rápido y capaz a menor costo",
"anthropic_claude_sonnet_4_6": "Rendimiento equilibrado",
"anthropic_claude_sonnet_4_5": "Generación anterior de Sonnet",
@@ -2616,6 +2617,7 @@
"groq_openai_gpt_oss_20b": "Modelo de código abierto rápido, 1000 T/seg",
"groq_compound": "Sistema compound de Groq, 450 T/seg",
"groq_compound_mini": "Sistema compound rápido, 3x menor latencia",
+ "openai_gpt_6_astra": "El modelo más capaz de OpenAI para programación e investigación",
"openai_gpt_5_6_sol": "GPT-5.6 insignia para programación e investigación complejas",
"openai_gpt_5_6_terra": "GPT-5.6 equilibrado para el trabajo diario",
"openai_gpt_5_6_luna": "El GPT-5.6 más rápido y económico",
diff --git a/src/locales/fr/translation.json b/src/locales/fr/translation.json
index 940573987..8cf6e3026 100644
--- a/src/locales/fr/translation.json
+++ b/src/locales/fr/translation.json
@@ -2644,7 +2644,8 @@
"openai_gpt_4_1": "Base solide, contexte 1M",
"openai_gpt_4_1_mini": "Modèle GPT-4.1 plus petit",
"openai_gpt_4_1_nano": "GPT-4.1 à la latence la plus faible",
- "anthropic_claude_fable_5": "Le modèle le plus performant d'Anthropic, classe Mythos",
+ "anthropic_claude_fable_5_1": "Le modèle le plus performant d'Anthropic, classe Mythos",
+ "anthropic_claude_fable_5": "Ancien modèle phare de classe Mythos",
"anthropic_claude_sonnet_5": "Modèle assistant rapide et performant à moindre coût",
"anthropic_claude_sonnet_4_6": "Performances équilibrées",
"anthropic_claude_sonnet_4_5": "Génération Sonnet précédente",
@@ -2661,6 +2662,7 @@
"groq_openai_gpt_oss_20b": "Modèle open source rapide, 1000 T/s",
"groq_compound": "Système compound de Groq, 450 T/s",
"groq_compound_mini": "Système compound rapide, 3x latence réduite",
+ "openai_gpt_6_astra": "Le modèle le plus performant d'OpenAI pour le codage et la recherche",
"openai_gpt_5_6_sol": "GPT-5.6 phare pour le codage et la recherche complexes",
"openai_gpt_5_6_terra": "GPT-5.6 équilibré pour le travail quotidien",
"openai_gpt_5_6_luna": "Le GPT-5.6 le plus rapide et le plus économique",
diff --git a/src/locales/it/translation.json b/src/locales/it/translation.json
index fc9fd2191..9192c4c8e 100644
--- a/src/locales/it/translation.json
+++ b/src/locales/it/translation.json
@@ -2551,7 +2551,8 @@
"openai_gpt_4_1": "Base solida, contesto da 1M",
"openai_gpt_4_1_mini": "Modello GPT-4.1 più piccolo",
"openai_gpt_4_1_nano": "GPT-4.1 con la latenza più bassa",
- "anthropic_claude_fable_5": "Il modello più capace di Anthropic, classe Mythos",
+ "anthropic_claude_fable_5_1": "Il modello più capace di Anthropic, classe Mythos",
+ "anthropic_claude_fable_5": "Precedente modello di punta di classe Mythos",
"anthropic_claude_sonnet_5": "Modello assistente veloce e capace a costo inferiore",
"anthropic_claude_sonnet_4_6": "Prestazioni bilanciate",
"anthropic_claude_sonnet_4_5": "Generazione Sonnet precedente",
@@ -2568,6 +2569,7 @@
"groq_openai_gpt_oss_20b": "Modello open source veloce, 1000 T/sec",
"groq_compound": "Sistema compound di Groq, 450 T/sec",
"groq_compound_mini": "Sistema compound veloce, 3x latenza ridotta",
+ "openai_gpt_6_astra": "Il modello più capace di OpenAI per programmazione e ricerca",
"openai_gpt_5_6_sol": "GPT-5.6 di punta per programmazione e ricerca complesse",
"openai_gpt_5_6_terra": "GPT-5.6 bilanciato per il lavoro quotidiano",
"openai_gpt_5_6_luna": "Il GPT-5.6 più veloce ed economico",
diff --git a/src/locales/ja/translation.json b/src/locales/ja/translation.json
index ecb4c2278..5ee6c168d 100644
--- a/src/locales/ja/translation.json
+++ b/src/locales/ja/translation.json
@@ -2551,7 +2551,8 @@
"openai_gpt_4_1": "強力なベースライン、100 万トークンコンテキスト",
"openai_gpt_4_1_mini": "小型 GPT-4.1 モデル",
"openai_gpt_4_1_nano": "最低レイテンシーの GPT-4.1",
- "anthropic_claude_fable_5": "Anthropic の最も高性能なモデル、Mythos クラス",
+ "anthropic_claude_fable_5_1": "Anthropic の最も高性能なモデル、Mythos クラス",
+ "anthropic_claude_fable_5": "前世代の Mythos クラスのフラッグシップ",
"anthropic_claude_sonnet_5": "低コストで高速かつ高性能なアシスタントモデル",
"anthropic_claude_sonnet_4_6": "バランスの取れた性能",
"anthropic_claude_sonnet_4_5": "前世代の Sonnet モデル",
@@ -2568,6 +2569,7 @@
"groq_openai_gpt_oss_20b": "高速オープンソースモデル、1000 T/秒",
"groq_compound": "Groq の Compound システム、450 T/秒",
"groq_compound_mini": "高速 Compound システム、3x 低レイテンシー",
+ "openai_gpt_6_astra": "コーディングと研究のための OpenAI の最も高性能なモデル",
"openai_gpt_5_6_sol": "複雑なコーディングと研究のためのフラッグシップ GPT-5.6",
"openai_gpt_5_6_terra": "日常業務向けのバランス型 GPT-5.6",
"openai_gpt_5_6_luna": "最速かつ最も低コストの GPT-5.6",
diff --git a/src/locales/pt/translation.json b/src/locales/pt/translation.json
index 7611adee6..cca76eebd 100644
--- a/src/locales/pt/translation.json
+++ b/src/locales/pt/translation.json
@@ -2530,7 +2530,8 @@
"openai_gpt_4_1": "Base sólida, contexto de 1M",
"openai_gpt_4_1_mini": "Modelo GPT-4.1 menor",
"openai_gpt_4_1_nano": "GPT-4.1 com menor latência",
- "anthropic_claude_fable_5": "O modelo mais capaz da Anthropic, classe Mythos",
+ "anthropic_claude_fable_5_1": "O modelo mais capaz da Anthropic, classe Mythos",
+ "anthropic_claude_fable_5": "Modelo principal anterior da classe Mythos",
"anthropic_claude_sonnet_5": "Modelo assistente rápido e capaz por menor custo",
"anthropic_claude_sonnet_4_6": "Desempenho equilibrado",
"anthropic_claude_sonnet_4_5": "Geração anterior do Sonnet",
@@ -2547,6 +2548,7 @@
"groq_openai_gpt_oss_20b": "Modelo open source rápido, 1000 T/seg",
"groq_compound": "Sistema compound da Groq, 450 T/seg",
"groq_compound_mini": "Sistema compound rápido, 3x menor latência",
+ "openai_gpt_6_astra": "O modelo mais capaz da OpenAI para programação e pesquisa",
"openai_gpt_5_6_sol": "GPT-5.6 principal para programação e pesquisa complexas",
"openai_gpt_5_6_terra": "GPT-5.6 equilibrado para o trabalho do dia a dia",
"openai_gpt_5_6_luna": "O GPT-5.6 mais rápido e econômico",
diff --git a/src/locales/ru/translation.json b/src/locales/ru/translation.json
index 22fb44608..4e44f69fd 100644
--- a/src/locales/ru/translation.json
+++ b/src/locales/ru/translation.json
@@ -2567,7 +2567,8 @@
"openai_gpt_4_1": "Надёжная базовая модель, контекст 1M",
"openai_gpt_4_1_mini": "Компактная версия GPT-4.1",
"openai_gpt_4_1_nano": "Минимальная задержка GPT-4.1",
- "anthropic_claude_fable_5": "Самая мощная модель Anthropic, класс Mythos",
+ "anthropic_claude_fable_5_1": "Самая мощная модель Anthropic, класс Mythos",
+ "anthropic_claude_fable_5": "Предыдущая флагманская модель класса Mythos",
"anthropic_claude_sonnet_5": "Быстрая и мощная модель-помощник по низкой цене",
"anthropic_claude_sonnet_4_6": "Сбалансированная производительность",
"anthropic_claude_sonnet_4_5": "Предыдущее поколение Sonnet",
@@ -2584,6 +2585,7 @@
"groq_openai_gpt_oss_20b": "Быстрая открытая модель, 1000 т/с",
"groq_compound": "Система Compound от Groq, 450 т/с",
"groq_compound_mini": "Быстрая система Compound, 3x ниже задержка",
+ "openai_gpt_6_astra": "Самая мощная модель OpenAI для программирования и исследований",
"openai_gpt_5_6_sol": "Флагманская GPT-5.6 для сложного программирования и исследований",
"openai_gpt_5_6_terra": "Сбалансированная GPT-5.6 для повседневной работы",
"openai_gpt_5_6_luna": "Самая быстрая и недорогая GPT-5.6",
diff --git a/src/locales/zh-CN/translation.json b/src/locales/zh-CN/translation.json
index 414cde9d8..8ac971500 100644
--- a/src/locales/zh-CN/translation.json
+++ b/src/locales/zh-CN/translation.json
@@ -2551,7 +2551,8 @@
"openai_gpt_4_1": "稳定基准,1M 上下文",
"openai_gpt_4_1_mini": "精简版 GPT-4.1",
"openai_gpt_4_1_nano": "最低延迟 GPT-4.1",
- "anthropic_claude_fable_5": "Anthropic 最强大的模型,Mythos 级",
+ "anthropic_claude_fable_5_1": "Anthropic 最强大的模型,Mythos 级",
+ "anthropic_claude_fable_5": "上一代 Mythos 级旗舰模型",
"anthropic_claude_sonnet_5": "快速且强大的助手模型,成本更低",
"anthropic_claude_sonnet_4_6": "均衡性能",
"anthropic_claude_sonnet_4_5": "上一代 Sonnet",
@@ -2568,6 +2569,7 @@
"groq_openai_gpt_oss_20b": "快速开源模型,1000 T/秒",
"groq_compound": "Groq Compound 系统,450 T/秒",
"groq_compound_mini": "快速 Compound 系统,3x 更低延迟",
+ "openai_gpt_6_astra": "OpenAI 最强大的模型,适用于编码与研究",
"openai_gpt_5_6_sol": "旗舰 GPT-5.6,适用于复杂编码与研究",
"openai_gpt_5_6_terra": "均衡的 GPT-5.6,适合日常工作",
"openai_gpt_5_6_luna": "速度最快、成本最低的 GPT-5.6",
diff --git a/src/locales/zh-TW/translation.json b/src/locales/zh-TW/translation.json
index a16525ee0..955b2dcb1 100644
--- a/src/locales/zh-TW/translation.json
+++ b/src/locales/zh-TW/translation.json
@@ -2551,7 +2551,8 @@
"openai_gpt_4_1": "穩健基礎,1M 上下文",
"openai_gpt_4_1_mini": "較小的 GPT-4.1 模型",
"openai_gpt_4_1_nano": "最低延遲的 GPT-4.1",
- "anthropic_claude_fable_5": "Anthropic 最強大的模型,Mythos 級",
+ "anthropic_claude_fable_5_1": "Anthropic 最強大的模型,Mythos 級",
+ "anthropic_claude_fable_5": "上一代 Mythos 級旗艦模型",
"anthropic_claude_sonnet_5": "快速且強大的助理模型,成本更低",
"anthropic_claude_sonnet_4_6": "均衡表現",
"anthropic_claude_sonnet_4_5": "上一代 Sonnet 模型",
@@ -2568,6 +2569,7 @@
"groq_openai_gpt_oss_20b": "快速開源模型,1000 T/sec",
"groq_compound": "Groq Compound 系統,450 T/sec",
"groq_compound_mini": "快速 Compound 系統,3x 更低延遲",
+ "openai_gpt_6_astra": "OpenAI 最強大的模型,適用於編碼與研究",
"openai_gpt_5_6_sol": "旗艦 GPT-5.6,適用於複雜編碼與研究",
"openai_gpt_5_6_terra": "均衡的 GPT-5.6,適合日常工作",
"openai_gpt_5_6_luna": "速度最快、成本最低的 GPT-5.6",
diff --git a/src/models/modelRegistryData.json b/src/models/modelRegistryData.json
index 616f2cd61..a81103d74 100644
--- a/src/models/modelRegistryData.json
+++ b/src/models/modelRegistryData.json
@@ -371,6 +371,15 @@
"id": "openai",
"name": "OpenAI",
"models": [
+ {
+ "id": "gpt-6-astra",
+ "name": "GPT-6 Astra",
+ "description": "OpenAI's most capable model for coding, computer use, and research, 1M context",
+ "descriptionKey": "models.descriptions.cloud.openai_gpt_6_astra",
+ "tokenParam": "max_completion_tokens",
+ "supportsTemperature": false,
+ "supportsVision": true
+ },
{
"id": "gpt-5.6-sol",
"name": "GPT-5.6 Sol",
@@ -467,10 +476,18 @@
"id": "anthropic",
"name": "Anthropic",
"models": [
+ {
+ "id": "claude-fable-5-1",
+ "name": "Claude Fable 5.1",
+ "description": "Anthropic's most capable model, Mythos-class, 1M context",
+ "descriptionKey": "models.descriptions.cloud.anthropic_claude_fable_5_1",
+ "supportsTemperature": false,
+ "supportsVision": true
+ },
{
"id": "claude-fable-5",
"name": "Claude Fable 5",
- "description": "Anthropic's most capable model, Mythos-class, 1M context",
+ "description": "Previous Mythos-class flagship, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_fable_5",
"supportsTemperature": false,
"supportsVision": true
diff --git a/test/helpers/anthropicSamplingFlags.test.js b/test/helpers/anthropicSamplingFlags.test.js
index 7f14a75c3..b8cff61df 100644
--- a/test/helpers/anthropicSamplingFlags.test.js
+++ b/test/helpers/anthropicSamplingFlags.test.js
@@ -17,6 +17,7 @@ test("every Anthropic model declares sampling support explicitly", () => {
test("models from Opus 4.7 onward do not receive temperature", () => {
for (const id of [
+ "claude-fable-5-1",
"claude-fable-5",
"claude-sonnet-5",
"claude-opus-5",