torbjorn/my-pal-mcp-server
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs/openrouter-sync-operations #1

Merged
torbjorn merged 6 commits from docs/openrouter-sync-operations into main 2026-04-01 22:01:02 +00:00
Conversation 0 Commits 6 Files Changed 14 +1300 -9
7 changed files with 1300 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 7ef476cfbd - Show all commits

696
conf/zen_models_live.json Normal file
View File

@@ -0,0 +1,696 @@
{
"_README": {
"description": "Generated baseline OpenCode Zen catalogue for PAL MCP Server.",
"source": "https://opencode.ai/zen/v1/models",
"usage": "Generated by scripts/sync_zen_models.py. Curated overrides belong in conf/zen_models.json.",
"field_notes": "Entries are conservative discovery data. Curated manifest values override these at runtime."
},
"models": [
{
"model_name": "big-pickle",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 32000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Big Pickle via OpenCode Zen - Stealth model for coding tasks",
"intelligence_score": 13,
"allow_code_generation": true
},
{
"model_name": "claude-3-5-haiku",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model claude-3-5-haiku.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "claude-haiku-4-5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 5.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Claude Haiku 4.5 via OpenCode Zen - Fast and efficient for coding tasks",
"intelligence_score": 16,
"allow_code_generation": true
},
{
"model_name": "claude-opus-4-1",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model claude-opus-4-1.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "claude-opus-4-5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 5.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Claude Opus 4.5 via OpenCode Zen - Anthropic's frontier reasoning model for complex software engineering",
"intelligence_score": 18,
"allow_code_generation": true
},
{
"model_name": "claude-opus-4-6",
"aliases": [],
"context_window": 1000000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model claude-opus-4-6.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "claude-sonnet-4",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model claude-sonnet-4.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "claude-sonnet-4-5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 5.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Claude Sonnet 4.5 via OpenCode Zen - Balanced performance for coding and general tasks",
"intelligence_score": 17,
"allow_code_generation": true
},
{
"model_name": "claude-sonnet-4-6",
"aliases": [],
"context_window": 1000000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model claude-sonnet-4-6.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "gemini-3-flash",
"aliases": [],
"context_window": 1048576,
"max_output_tokens": 65536,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gemini-3-flash.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "gemini-3-pro",
"aliases": [],
"context_window": 1000000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 10.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Gemini 3 Pro via OpenCode Zen - Google's multimodal model with large context",
"intelligence_score": 16,
"allow_code_generation": true
},
{
"model_name": "gemini-3.1-pro",
"aliases": [],
"context_window": 1048576,
"max_output_tokens": 65536,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gemini-3.1-pro.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "glm-4.6",
"aliases": [],
"context_window": 205000,
"max_output_tokens": 32000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "GLM 4.6 via OpenCode Zen - High-performance model for coding and reasoning",
"intelligence_score": 15,
"allow_code_generation": true
},
{
"model_name": "glm-4.7",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model glm-4.7.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "glm-5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model glm-5.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "gpt-5",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5-codex",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5-codex.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5-nano",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 32000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "GPT 5 Nano via OpenCode Zen - Lightweight GPT model",
"intelligence_score": 12,
"allow_code_generation": true
},
{
"model_name": "gpt-5.1",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "GPT 5.1 via OpenCode Zen - Latest GPT model for general AI tasks",
"intelligence_score": 16,
"allow_code_generation": true,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.1-codex",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "GPT 5.1 Codex via OpenCode Zen - Specialized for code generation and understanding",
"intelligence_score": 17,
"allow_code_generation": true,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.1-codex-max",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.1-codex-max.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.1-codex-mini",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.1-codex-mini.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.2",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.2.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.2-codex",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.2-codex.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.3-codex",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.3-codex.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.3-codex-spark",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.3-codex-spark.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.4",
"aliases": [],
"context_window": 1050000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.4.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.4-mini",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.4-mini.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.4-nano",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.4-nano.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "gpt-5.4-pro",
"aliases": [],
"context_window": 1050000,
"max_output_tokens": 128000,
"supports_extended_thinking": true,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": true,
"max_image_size_mb": 20.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model gpt-5.4-pro.",
"intelligence_score": 10,
"allow_code_generation": false,
"use_openai_response_api": true
},
{
"model_name": "kimi-k2",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 32000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Kimi K2 via OpenCode Zen - Advanced reasoning model",
"intelligence_score": 15,
"allow_code_generation": true
},
{
"model_name": "kimi-k2-thinking",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model kimi-k2-thinking.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "kimi-k2.5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model kimi-k2.5.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "mimo-v2-flash-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model mimo-v2-flash-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "mimo-v2-omni-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model mimo-v2-omni-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "mimo-v2-pro-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model mimo-v2-pro-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "minimax-m2.1",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model minimax-m2.1.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "minimax-m2.5",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model minimax-m2.5.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "minimax-m2.5-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model minimax-m2.5-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "nemotron-3-super-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model nemotron-3-super-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "qwen3.6-plus-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model qwen3.6-plus-free.",
"intelligence_score": 10,
"allow_code_generation": false
},
{
"model_name": "trinity-large-preview-free",
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": false,
"supports_json_mode": true,
"supports_function_calling": true,
"supports_images": false,
"max_image_size_mb": 0.0,
"supports_temperature": true,
"temperature_constraint": "range",
"description": "Generated baseline metadata for OpenCode Zen model trinity-large-preview-free.",
"intelligence_score": 10,
"allow_code_generation": false
}
]
}

5
docs/configuration.md
View File

@@ -79,6 +79,8 @@ DEFAULT_MODEL=auto # Claude picks best model for each task (recommended)
- `conf/openai_models.json` – OpenAI catalogue (can be overridden with `OPENAI_MODELS_CONFIG_PATH`)
- `conf/gemini_models.json` – Gemini catalogue (`GEMINI_MODELS_CONFIG_PATH`)
- `conf/xai_models.json` – X.AI / GROK catalogue (`XAI_MODELS_CONFIG_PATH`)
- `conf/zen_models.json` – Curated OpenCode Zen overrides (`ZEN_MODELS_CONFIG_PATH`)
- `conf/zen_models_live.json` – Generated live OpenCode Zen catalogue (`ZEN_LIVE_MODELS_CONFIG_PATH`)
- `conf/openrouter_models.json` – Curated OpenRouter overrides (`OPENROUTER_MODELS_CONFIG_PATH`)
- `conf/openrouter_models_live.json` – Generated live OpenRouter catalogue (`OPENROUTER_LIVE_MODELS_CONFIG_PATH`)
- `conf/dial_models.json` – DIAL aggregation catalogue (`DIAL_MODELS_CONFIG_PATH`)
@@ -93,10 +95,11 @@ DEFAULT_MODEL=auto # Claude picks best model for each task (recommended)
| OpenAI | `gpt-5.2`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, `gpt-5`, `gpt-5.2-pro`, `gpt-5-mini`, `gpt-5-nano`, `gpt-5-codex`, `gpt-4.1`, `o3`, `o3-mini`, `o3-pro`, `o4-mini` | `gpt5.2`, `gpt-5.2`, `5.2`, `gpt5.1-codex`, `codex-5.1`, `codex-mini`, `gpt5`, `gpt5pro`, `mini`, `nano`, `codex`, `o3mini`, `o3pro`, `o4mini` |
| Gemini | `gemini-2.5-pro`, `gemini-2.5-flash`, `gemini-2.0-flash`, `gemini-2.0-flash-lite` | `pro`, `gemini-pro`, `flash`, `flash-2.0`, `flashlite` |
| X.AI | `grok-4`, `grok-4.1-fast` | `grok`, `grok4`, `grok-4.1-fast-reasoning` |
| OpenCode Zen | Generated live catalogue plus curated overrides | e.g., `zen-sonnet`, `zen-codex`, plus any curated aliases you add |
| OpenRouter | Generated live catalogue plus curated overrides | e.g., `opus`, `sonnet`, `flash`, `pro`, `mistral` |
| Custom | User-managed entries such as `llama3.2` | Define your own aliases per entry |
Latest OpenAI entries (`gpt-5.2`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, `gpt-5.2-pro`) expose 400K-token contexts with large outputs, reasoning-token support, and multimodal inputs. `gpt-5.1-codex` and `gpt-5.2-pro` are Responses-only with streaming disabled, while the base `gpt-5.2` and Codex mini support streaming along with full code-generation flags. For OpenRouter, keep PAL-specific metadata in the curated manifest and regenerate the live catalogue when OpenRouter adds or removes models; see [Refreshing the Live OpenRouter Catalogue](custom_models.md#refreshing-the-live-openrouter-catalogue).
Latest OpenAI entries (`gpt-5.2`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, `gpt-5.2-pro`) expose 400K-token contexts with large outputs, reasoning-token support, and multimodal inputs. `gpt-5.1-codex` and `gpt-5.2-pro` are Responses-only with streaming disabled, while the base `gpt-5.2` and Codex mini support streaming along with full code-generation flags. For OpenCode Zen and OpenRouter, keep PAL-specific metadata in the curated manifest and regenerate the live catalogue when the upstream provider adds or removes models; see [Refreshing the Live OpenCode Zen Catalogue](custom_models.md#refreshing-the-live-opencode-zen-catalogue) and [Refreshing the Live OpenRouter Catalogue](custom_models.md#refreshing-the-live-openrouter-catalogue).
> **Tip:** Copy the JSON file you need, customise it, and point the corresponding `*_MODELS_CONFIG_PATH` environment variable to your version. This lets you enable or disable capabilities (JSON mode, function calling, temperature support, code generation) without editing Python.

47
docs/custom_models.md
View File

@@ -221,13 +221,15 @@ CUSTOM_MODEL_NAME=your-loaded-model
The system automatically routes models to the appropriate provider:
1. Entries in `conf/custom_models.json` → Always routed through the Custom API (requires `CUSTOM_API_URL`)
2. Entries in `conf/openrouter_models_live.json` and `conf/openrouter_models.json` → Routed through OpenRouter (requires `OPENROUTER_API_KEY`)
3. **Unknown models** → Fallback logic based on model name patterns
2. Entries in `conf/zen_models_live.json` and `conf/zen_models.json` → Routed through OpenCode Zen (requires `ZEN_API_KEY`)
3. Entries in `conf/openrouter_models_live.json` and `conf/openrouter_models.json` → Routed through OpenRouter (requires `OPENROUTER_API_KEY`)
4. **Unknown models** → Fallback logic based on model name patterns
**Provider Priority Order:**
1. Native APIs (Google, OpenAI) - if API keys are available
2. Custom endpoints - for models declared in `conf/custom_models.json`
3. OpenRouter - catch-all for cloud models
2. OpenCode Zen - curated gateway models
3. Custom endpoints - for models declared in `conf/custom_models.json`
4. OpenRouter - catch-all for cloud models
This ensures clean separation between local and cloud models while maintaining flexibility for unknown models.
@@ -241,7 +243,42 @@ These JSON files define model aliases and capabilities. You can:
### Adding Custom Models
Edit `conf/openrouter_models.json` to tweak OpenRouter behaviour or `conf/custom_models.json` to add local models. The generated `conf/openrouter_models_live.json` file is discovery data from OpenRouter's `/api/v1/models` endpoint; curated entries in `conf/openrouter_models.json` override those generated defaults. Each entry maps directly onto [`ModelCapabilities`](../providers/shared/model_capabilities.py).
Edit `conf/zen_models.json` to tweak OpenCode Zen behaviour, `conf/openrouter_models.json` to tweak OpenRouter behaviour, or `conf/custom_models.json` to add local models. The generated `conf/zen_models_live.json` and `conf/openrouter_models_live.json` files are discovery data from the upstream model listing endpoints; curated entries override those generated defaults. Each entry maps directly onto [`ModelCapabilities`](../providers/shared/model_capabilities.py).
### Refreshing the Live OpenCode Zen Catalogue
Run the sync script whenever OpenCode Zen adds or removes models that you want `listmodels` and provider enumeration to expose, or before cutting a release that should include an updated Zen catalogue.
```bash
source .pal_venv/bin/activate
python scripts/sync_zen_models.py
```
By default the script:
- fetches `https://opencode.ai/zen/v1/models`
- writes conservative discovery data to `conf/zen_models_live.json`
- leaves `conf/zen_models.json` untouched
Use the optional flags if you need to test against a different endpoint or write to a different file:
```bash
python scripts/sync_zen_models.py --url https://opencode.ai/zen/v1/models --output conf/zen_models_live.json
```
Important runtime behavior:
- `conf/zen_models_live.json` is the generated baseline catalogue
- `conf/zen_models.json` is the curated override layer for aliases and PAL-specific capability flags
- curated entries win when the same `model_name` appears in both files
- models missing from the curated file are still available from the generated catalogue
After refreshing the catalogue:
1. Review the diff in `conf/zen_models_live.json`
2. Add or update curated entries in `conf/zen_models.json` if a new model needs aliases or PAL-specific capability tweaks
3. Restart the server so the updated manifests are reloaded
4. Commit the generated JSON alongside any curated overrides so other contributors get the same catalogue state
### Refreshing the Live OpenRouter Catalogue

90
providers/registries/zen.py
View File

@@ -2,14 +2,48 @@
from __future__ import annotations
import importlib.resources
import json
import logging
from pathlib import Path
from utils.env import get_env
from utils.file_utils import read_json_file
from ..shared import ModelCapabilities, ProviderType
from .base import CAPABILITY_FIELD_NAMES, CapabilityModelRegistry
logger = logging.getLogger(__name__)
class ZenModelRegistry(CapabilityModelRegistry):
"""Capability registry backed by ``conf/zen_models.json``."""
def __init__(self, config_path: str | None = None) -> None:
LIVE_ENV_VAR_NAME = "ZEN_LIVE_MODELS_CONFIG_PATH"
LIVE_DEFAULT_FILENAME = "zen_models_live.json"
def __init__(self, config_path: str | None = None, live_config_path: str | None = None) -> None:
self._live_resource = None
self._live_config_path: Path | None = None
self._live_default_path = Path(__file__).resolve().parents[3] / "conf" / self.LIVE_DEFAULT_FILENAME
if live_config_path:
self._live_config_path = Path(live_config_path)
else:
env_path = get_env(self.LIVE_ENV_VAR_NAME)
if env_path:
self._live_config_path = Path(env_path)
else:
try:
resource = importlib.resources.files("conf").joinpath(self.LIVE_DEFAULT_FILENAME)
if hasattr(resource, "read_text"):
self._live_resource = resource
else:
raise AttributeError("resource accessor not available")
except Exception:
self._live_config_path = self._live_default_path
super().__init__(
env_var_name="ZEN_MODELS_CONFIG_PATH",
default_filename="zen_models.json",
@@ -18,6 +52,60 @@ class ZenModelRegistry(CapabilityModelRegistry):
config_path=config_path,
)
def reload(self) -> None:
live_data = self._load_live_config_data()
curated_data = self._load_config_data()
merged_data = self._merge_manifest_data(live_data, curated_data)
self._extras = {}
configs = [config for config in self._parse_models(merged_data) if config is not None]
self._build_maps(configs)
def _load_live_config_data(self) -> dict:
if self._live_resource is not None:
try:
if hasattr(self._live_resource, "read_text"):
config_text = self._live_resource.read_text(encoding="utf-8")
else:
with self._live_resource.open("r", encoding="utf-8") as handle:
config_text = handle.read()
data = json.loads(config_text)
except FileNotFoundError:
logger.debug("Packaged %s not found", self.LIVE_DEFAULT_FILENAME)
return {"models": []}
except Exception as exc:
logger.warning("Failed to read packaged %s: %s", self.LIVE_DEFAULT_FILENAME, exc)
return {"models": []}
return data or {"models": []}
if not self._live_config_path:
return {"models": []}
if not self._live_config_path.exists():
logger.debug("Zen live registry config not found at %s", self._live_config_path)
return {"models": []}
data = read_json_file(str(self._live_config_path))
return data or {"models": []}
@staticmethod
def _merge_manifest_data(live_data: dict, curated_data: dict) -> dict:
merged_models: dict[str, dict] = {}
for source in (live_data, curated_data):
for raw in source.get("models", []):
if not isinstance(raw, dict):
continue
model_name = raw.get("model_name")
if not model_name:
continue
existing = merged_models.get(model_name, {})
merged_models[model_name] = {**existing, **dict(raw)}
return {"models": list(merged_models.values())}
def _finalise_entry(self, entry: dict) -> tuple[ModelCapabilities, dict]:
provider_override = entry.get("provider")
if isinstance(provider_override, str):

237
scripts/sync_zen_models.py Normal file
View File

@@ -0,0 +1,237 @@
#!/usr/bin/env python3
import argparse
import json
import sys
from pathlib import Path
from urllib.error import HTTPError, URLError
from urllib.request import Request, urlopen
from utils.env import get_env
from utils.file_utils import read_json_file
ROOT = Path(__file__).resolve().parents[1]
DEFAULT_OUTPUT = ROOT / "conf" / "zen_models_live.json"
DEFAULT_CURATED = ROOT / "conf" / "zen_models.json"
ZEN_MODELS_URL = "https://opencode.ai/zen/v1/models"
def fetch_zen_models(url: str, api_key: str) -> dict:
request = Request(
url,
headers={
"Accept": "application/json",
"Authorization": f"Bearer {api_key}",
"User-Agent": "pal-mcp-server/zen-model-sync",
},
)
with urlopen(request, timeout=30) as response:
charset = response.headers.get_content_charset("utf-8")
payload = response.read().decode(charset)
data = json.loads(payload)
if not isinstance(data, dict):
raise ValueError("Zen models payload must be a JSON object")
return data
def load_curated_models(path: Path) -> dict[str, dict]:
if not path.exists():
return {}
data = read_json_file(str(path)) or {}
models = data.get("models", [])
if not isinstance(models, list):
return {}
curated: dict[str, dict] = {}
for item in models:
if not isinstance(item, dict):
continue
model_name = item.get("model_name")
if isinstance(model_name, str) and model_name:
curated[model_name] = dict(item)
return curated
def _infer_defaults_from_model_name(model_name: str) -> dict:
lower_name = model_name.lower()
defaults = {
"aliases": [],
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": False,
"supports_json_mode": True,
"supports_function_calling": True,
"supports_images": False,
"max_image_size_mb": 0.0,
"supports_temperature": True,
"temperature_constraint": "range",
"description": f"OpenCode Zen live model: {model_name}",
"intelligence_score": 10,
"allow_code_generation": False,
}
if lower_name.startswith("claude-"):
defaults.update(
{
"supports_extended_thinking": True,
"supports_images": True,
"max_image_size_mb": 20.0,
}
)
if "-4-6" in lower_name:
defaults.update({"context_window": 1000000, "max_output_tokens": 128000})
elif lower_name.startswith("gemini-"):
defaults.update(
{
"context_window": 1048576,
"max_output_tokens": 65536,
"supports_extended_thinking": True,
"supports_images": True,
"max_image_size_mb": 20.0,
}
)
elif lower_name.startswith("gpt-"):
defaults.update(
{
"context_window": 400000,
"max_output_tokens": 128000,
"supports_extended_thinking": True,
"supports_images": True,
"max_image_size_mb": 20.0,
}
)
if "5.4" in lower_name:
defaults["context_window"] = 1050000 if "-pro" in lower_name or lower_name == "gpt-5.4" else 400000
if lower_name in {
"gpt-5.4",
"gpt-5.4-pro",
"gpt-5.4-mini",
"gpt-5.4-nano",
"gpt-5.3-codex",
"gpt-5.3-codex-spark",
"gpt-5.2",
"gpt-5.2-codex",
"gpt-5.1",
"gpt-5.1-codex",
"gpt-5.1-codex-max",
"gpt-5.1-codex-mini",
"gpt-5",
"gpt-5-codex",
}:
defaults["use_openai_response_api"] = True
return defaults
def convert_model(model_data: dict, curated_models: dict[str, dict]) -> dict | None:
model_name = model_data.get("id")
if not isinstance(model_name, str) or not model_name:
return None
curated_entry = curated_models.get(model_name, {})
defaults = _infer_defaults_from_model_name(model_name)
return {
"model_name": model_name,
"aliases": [],
"context_window": int(curated_entry.get("context_window", defaults["context_window"])),
"max_output_tokens": int(curated_entry.get("max_output_tokens", defaults["max_output_tokens"])),
"supports_extended_thinking": bool(
curated_entry.get("supports_extended_thinking", defaults["supports_extended_thinking"])
),
"supports_json_mode": bool(curated_entry.get("supports_json_mode", defaults["supports_json_mode"])),
"supports_function_calling": bool(
curated_entry.get("supports_function_calling", defaults["supports_function_calling"])
),
"supports_images": bool(curated_entry.get("supports_images", defaults["supports_images"])),
"max_image_size_mb": float(curated_entry.get("max_image_size_mb", defaults["max_image_size_mb"])),
"supports_temperature": bool(curated_entry.get("supports_temperature", defaults["supports_temperature"])),
"temperature_constraint": curated_entry.get("temperature_constraint", defaults["temperature_constraint"]),
"description": curated_entry.get("description")
or f"Generated baseline metadata for OpenCode Zen model {model_name}.",
"intelligence_score": int(curated_entry.get("intelligence_score", defaults["intelligence_score"])),
"allow_code_generation": bool(curated_entry.get("allow_code_generation", defaults["allow_code_generation"])),
**(
{"use_openai_response_api": bool(curated_entry.get("use_openai_response_api", True))}
if curated_entry.get("use_openai_response_api") is not None or defaults.get("use_openai_response_api")
else {}
),
}
def build_output_document(source: dict, source_url: str, curated_models: dict[str, dict]) -> dict:
models = []
for model_data in source.get("data", []):
if not isinstance(model_data, dict):
continue
converted = convert_model(model_data, curated_models)
if converted:
models.append(converted)
models.sort(key=lambda item: item["model_name"])
return {
"_README": {
"description": "Generated baseline OpenCode Zen catalogue for PAL MCP Server.",
"source": source_url,
"usage": "Generated by scripts/sync_zen_models.py. Curated overrides belong in conf/zen_models.json.",
"field_notes": "Entries are conservative discovery data. Curated manifest values override these at runtime.",
},
"models": models,
}
def write_output(path: Path, document: dict) -> None:
path.parent.mkdir(parents=True, exist_ok=True)
with path.open("w", encoding="utf-8", newline="\n") as handle:
json.dump(document, handle, indent=2, ensure_ascii=False)
handle.write("\n")
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(description="Sync OpenCode Zen live model catalogue into PAL config.")
parser.add_argument("--url", default=ZEN_MODELS_URL, help="Zen models endpoint")
parser.add_argument(
"--output",
default=str(DEFAULT_OUTPUT),
help="Path to the generated live Zen manifest",
)
parser.add_argument(
"--curated",
default=str(DEFAULT_CURATED),
help="Path to the curated Zen manifest used for metadata enrichment",
)
return parser.parse_args()
def main() -> int:
args = parse_args()
output_path = Path(args.output)
curated_path = Path(args.curated)
api_key = get_env("ZEN_API_KEY")
if not api_key:
print("Failed to sync Zen models: ZEN_API_KEY is not set", file=sys.stderr)
return 1
try:
curated_models = load_curated_models(curated_path)
source = fetch_zen_models(args.url, api_key)
document = build_output_document(source, args.url, curated_models)
write_output(output_path, document)
except (HTTPError, URLError, TimeoutError, ValueError, json.JSONDecodeError) as exc:
print(f"Failed to sync Zen models: {exc}", file=sys.stderr)
return 1
print(f"Wrote {len(document['models'])} Zen models to {output_path}")
return 0
if __name__ == "__main__":
raise SystemExit(main())

61
tests/test_sync_zen_models.py Normal file
View File

@@ -0,0 +1,61 @@
import importlib.util
from pathlib import Path
SCRIPT_PATH = Path(__file__).resolve().parents[1] / "scripts" / "sync_zen_models.py"
SPEC = importlib.util.spec_from_file_location("sync_zen_models", SCRIPT_PATH)
assert SPEC is not None and SPEC.loader is not None
MODULE = importlib.util.module_from_spec(SPEC)
SPEC.loader.exec_module(MODULE)
def test_convert_model_applies_family_defaults_for_gpt_5_4():
converted = MODULE.convert_model({"id": "gpt-5.4"}, curated_models={})
assert converted is not None
assert converted["model_name"] == "gpt-5.4"
assert converted["context_window"] == 1050000
assert converted["max_output_tokens"] == 128000
assert converted["supports_extended_thinking"] is True
assert converted["supports_images"] is True
assert converted["use_openai_response_api"] is True
def test_convert_model_prefers_curated_metadata_when_available():
converted = MODULE.convert_model(
{"id": "claude-opus-4-5"},
curated_models={
"claude-opus-4-5": {
"context_window": 200000,
"max_output_tokens": 64000,
"supports_extended_thinking": False,
"description": "Curated Opus 4.5",
"intelligence_score": 18,
"allow_code_generation": True,
}
},
)
assert converted is not None
assert converted["model_name"] == "claude-opus-4-5"
assert converted["context_window"] == 200000
assert converted["max_output_tokens"] == 64000
assert converted["supports_extended_thinking"] is False
assert converted["description"] == "Curated Opus 4.5"
assert converted["allow_code_generation"] is True
def test_build_output_document_sorts_model_ids():
document = MODULE.build_output_document(
{
"data": [
{"id": "gpt-5.4-pro"},
{"id": "claude-opus-4-6"},
]
},
"https://opencode.ai/zen/v1/models",
curated_models={},
)
model_names = [item["model_name"] for item in document["models"]]
assert model_names == sorted(model_names)

173
tests/test_zen_model_registry.py
View File

@@ -3,6 +3,7 @@
import json
import os
import tempfile
from pathlib import Path
from unittest.mock import patch
from providers.registries.zen import ZenModelRegistry
@@ -43,14 +44,19 @@ class TestZenModelRegistry:
json.dump(config_data, f)
temp_path = f.name
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as live_file:
json.dump({"models": []}, live_file)
live_path = live_file.name
try:
registry = ZenModelRegistry(config_path=temp_path)
registry = ZenModelRegistry(config_path=temp_path, live_config_path=live_path)
assert len(registry.list_models()) == 1
assert "test/zen-model-1" in registry.list_models()
assert "zen-test1" in registry.list_aliases()
assert "zt1" in registry.list_aliases()
finally:
os.unlink(temp_path)
os.unlink(live_path)
def test_get_capabilities(self):
"""Test capability retrieval."""
@@ -158,9 +164,172 @@ class TestZenModelRegistry:
json.dump(empty_config, f)
temp_path = f.name
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as live_file:
json.dump({"models": []}, live_file)
live_path = live_file.name
try:
registry = ZenModelRegistry(config_path=temp_path)
registry = ZenModelRegistry(config_path=temp_path, live_config_path=live_path)
assert len(registry.list_models()) == 0
assert len(registry.list_aliases()) == 0
finally:
os.unlink(temp_path)
os.unlink(live_path)
def test_live_catalogue_adds_unsynced_model_ids(self):
curated_data = {
"models": [
{
"model_name": "gpt-5.1",
"aliases": ["zen-gpt5.1"],
"context_window": 400000,
"max_output_tokens": 64000,
"intelligence_score": 16,
}
]
}
live_data = {
"models": [
{
"model_name": "gpt-5.4",
"aliases": [],
"context_window": 1050000,
"max_output_tokens": 128000,
"supports_extended_thinking": True,
"supports_json_mode": True,
"supports_function_calling": True,
"supports_images": True,
"max_image_size_mb": 20.0,
"supports_temperature": True,
"temperature_constraint": "range",
"description": "Live-only GPT-5.4 entry",
"use_openai_response_api": True,
}
]
}
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as curated_file:
json.dump(curated_data, curated_file)
curated_path = curated_file.name
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as live_file:
json.dump(live_data, live_file)
live_path = live_file.name
try:
registry = ZenModelRegistry(config_path=curated_path, live_config_path=live_path)
assert "gpt-5.4" in registry.list_models()
caps = registry.resolve("gpt-5.4")
assert caps is not None
assert caps.description == "Live-only GPT-5.4 entry"
assert caps.use_openai_response_api is True
finally:
os.unlink(curated_path)
os.unlink(live_path)
def test_curated_manifest_overrides_live_metadata(self):
curated_data = {
"models": [
{
"model_name": "gpt-5.4",
"aliases": ["zen-gpt5.4"],
"context_window": 1050000,
"max_output_tokens": 128000,
"supports_extended_thinking": True,
"supports_json_mode": True,
"supports_function_calling": True,
"supports_images": True,
"max_image_size_mb": 20.0,
"supports_temperature": False,
"temperature_constraint": "fixed",
"description": "Curated override",
"intelligence_score": 19,
"allow_code_generation": True,
"use_openai_response_api": True,
}
]
}
live_data = {
"models": [
{
"model_name": "gpt-5.4",
"aliases": [],
"context_window": 400000,
"max_output_tokens": 64000,
"supports_extended_thinking": False,
"supports_json_mode": True,
"supports_function_calling": True,
"supports_images": False,
"max_image_size_mb": 0.0,
"supports_temperature": True,
"temperature_constraint": "range",
"description": "Live baseline",
"intelligence_score": 10,
"allow_code_generation": False,
}
]
}
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as curated_file:
json.dump(curated_data, curated_file)
curated_path = curated_file.name
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as live_file:
json.dump(live_data, live_file)
live_path = live_file.name
try:
registry = ZenModelRegistry(config_path=curated_path, live_config_path=live_path)
caps = registry.resolve("zen-gpt5.4")
assert caps is not None
assert caps.model_name == "gpt-5.4"
assert caps.description == "Curated override"
assert caps.context_window == 1050000
assert caps.max_output_tokens == 128000
assert caps.supports_images is True
assert caps.supports_temperature is False
assert caps.allow_code_generation is True
assert caps.use_openai_response_api is True
finally:
os.unlink(curated_path)
os.unlink(live_path)
def test_missing_live_catalogue_keeps_curated_models_working(self, monkeypatch):
missing_live_path = Path(tempfile.gettempdir()) / "pal-missing-zen-live.json"
if missing_live_path.exists():
missing_live_path.unlink()
monkeypatch.setenv("ZEN_LIVE_MODELS_CONFIG_PATH", str(missing_live_path))
registry = ZenModelRegistry()
assert "gpt-5.1" in registry.list_models()
assert registry.resolve("zen-gpt5.1") is not None
def test_invalid_live_json_keeps_curated_models_working(self):
curated_data = {
"models": [
{
"model_name": "gpt-5.1",
"aliases": ["zen-gpt5.1"],
"context_window": 400000,
"max_output_tokens": 64000,
"intelligence_score": 16,
}
]
}
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as curated_file:
json.dump(curated_data, curated_file)
curated_path = curated_file.name
with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as live_file:
live_file.write("{ invalid json }")
live_path = live_file.name
try:
registry = ZenModelRegistry(config_path=curated_path, live_config_path=live_path)
assert "gpt-5.1" in registry.list_models()
assert registry.resolve("zen-gpt5.1") is not None
finally:
os.unlink(curated_path)
os.unlink(live_path)