feat: Add 9 new models, remove 2 obsolete, update aliases by timeleft-- · Pull Request #411 · BeehiveInnovations/pal-mcp-server

timeleft-- · 2026-03-06T12:30:11Z

Summary

9 new models added: Claude Opus 4.6, Claude Sonnet 4.6, Gemini 3.1 Pro Preview, GPT-5.4 Pro, GPT-5.3 Codex, DeepSeek V3.2 Exp, Devstral 2512, Qwen 3.5 397B, MiniMax M2.5
2 obsolete models removed: meta-llama/llama-3-70b (8K context), perplexity/llama-3-sonar-large-32k-online (legacy)
Alias migrations: Generic aliases (opus, sonnet, pro, gpt5pro) moved to newest model versions. Version-specific aliases preserved for backward compatibility (opus4.5, sonnet4.5, gemini3.0, gpt5.2-pro)
Version bump: 9.8.2 → 9.9.0 (pyproject.toml, config.py, uv.lock)
All model IDs verified against OpenRouter API (/api/v1/models)

Details

New models

Model	Context	Score	Key capabilities
`anthropic/claude-opus-4.6`	1M	18	Vision
`anthropic/claude-sonnet-4.6`	1M	13	Vision
`google/gemini-3.1-pro-preview`	1M	19	Thinking, vision, function calling, code-gen
`openai/gpt-5.4-pro`	1.05M	19	Responses API, reasoning, code-gen
`openai/gpt-5.3-codex`	400K	19	Responses API, reasoning, code-gen
`deepseek/deepseek-v3.2-exp`	164K	16	Thinking
`mistralai/devstral-2512`	262K	15	Function calling, coding
`qwen/qwen3.5-397b-a17b`	262K	16	Thinking, vision, function calling
`minimax/minimax-m2.5`	197K	16	Function calling, SWE-Bench 80.2%

Alias migration (backward compatible)

Generic alias	Old target	New target	Version-specific alias
`opus`	claude-opus-4.5	claude-opus-4.6	`opus4.5` still works
`sonnet`	claude-sonnet-4.5	claude-sonnet-4.6	`sonnet4.5` still works
`pro`, `gemini`	gemini-3-pro-preview	gemini-3.1-pro-preview	`gemini3.0` still works
`gpt5pro`	gpt-5.2-pro	gpt-5.4-pro	`gpt5.2-pro` still works

Test plan

All 870 unit tests pass (14 skipped, 0 failures)
Alias resolution tests updated and cover all new models
Backward compatibility verified (version-specific aliases resolve correctly)
No duplicate aliases (76 total, case-insensitive uniqueness enforced)
Model IDs verified against live OpenRouter API

🤖 Generated with Claude Code

New models: Claude Opus 4.6, Claude Sonnet 4.6, Gemini 3.1 Pro, GPT-5.4 Pro, GPT-5.3 Codex, DeepSeek V3.2, Devstral 2512, Qwen 3.5 397B, MiniMax M2.5. Removed: meta-llama/llama-3-70b (8K context, obsolete), perplexity/llama-3-sonar-large-32k-online (legacy). Generic aliases (opus, sonnet, pro, gpt5pro) migrated to newest model versions. Version-specific aliases preserved for backward compatibility (opus4.5, sonnet4.5, gemini3.0, gpt5.2-pro). Version bumped to 9.9.0. All 870 tests pass. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

gemini-code-assist · 2026-03-06T12:32:27Z

Summary of Changes

Hello, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request updates the project's AI model registry to incorporate the latest offerings from OpenRouter. It introduces nine new, advanced models, removes two outdated ones, and intelligently reconfigures generic aliases to point to the newest versions while preserving older, version-specific aliases for seamless backward compatibility. The changes aim to enhance the system's capabilities with cutting-edge models and ensure the model configurations remain current and robust, supported by a minor version increment and thorough validation.

Highlights

New Model Additions: Added 9 new AI models, including Claude Opus 4.6, Claude Sonnet 4.6, Gemini 3.1 Pro Preview, GPT-5.4 Pro, GPT-5.3 Codex, DeepSeek V3.2 Exp, Devstral 2512, Qwen 3.5 397B, and MiniMax M2.5, significantly expanding the available model capabilities.
Obsolete Model Removals: Removed 2 obsolete models, meta-llama/llama-3-70b (8K context) and perplexity/llama-3-sonar-large-32k-online (legacy), to maintain an up-to-date and efficient model list.
Alias Migrations and Backward Compatibility: Updated generic aliases (e.g., 'opus', 'sonnet', 'pro', 'gpt5pro') to point to their newest model versions, while preserving version-specific aliases (e.g., 'opus4.5', 'sonnet4.5', 'gemini3.0', 'gpt5.2-pro') to ensure backward compatibility.
Version Bump: Bumped the project version from 9.8.2 to 9.9.0 across pyproject.toml, config.py, and uv.lock, indicating a minor feature release.
Model ID Verification: Verified all model IDs against the live OpenRouter API (/api/v1/models) to ensure accuracy and functionality of the updated configurations.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Changelog

conf/openrouter_models.json
- Added configurations for 9 new models including Claude Opus 4.6, Gemini 3.1 Pro Preview, and GPT-5.4 Pro.
- Removed configurations for meta-llama/llama-3-70b and perplexity/llama-3-sonar-large-32k-online.
- Updated aliases for existing models to shift generic aliases to newer versions while retaining version-specific ones.
config.py
- Updated the project version to 9.9.0.
- Updated the last update date to 2026-03-06.
docs/custom_models.md
- Updated the model alias table to reflect new generic alias mappings and the inclusion of new models.
- Removed documentation entries for deprecated models and their aliases.
pyproject.toml
- Updated the project version to 9.9.0.
- Updated the minimum required Python version to >=3.10.
tests/test_custom_provider.py
- Modified alias resolution tests to reflect the removal of the 'llama' alias and updated 'deepseek' alias behavior.
tests/test_model_resolution_bug.py
- Updated the Gemini alias resolution test to confirm generic 'gemini' and 'pro' aliases resolve to 'google/gemini-3.1-pro-preview', while 'gemini3.0' remains for backward compatibility.
tests/test_openrouter_provider.py
- Updated alias resolution tests to reflect new generic alias targets for 'opus', 'sonnet', 'pro', and 'gemini3.0'.
- Updated capability tests to reflect the 'opus' alias now pointing to 'anthropic/claude-opus-4.6' with its increased context window.
- Updated alias resolution for 'sonnet' to point to 'anthropic/claude-sonnet-4.6'.
tests/test_openrouter_registry.py
- Updated alias resolution test cases to include new alias mappings for 'opus', 'sonnet', 'pro', 'gemini3.1', 'gemini3.0', and new model aliases.
- Updated model capabilities conversion test to reflect 'opus' alias pointing to 'anthropic/claude-opus-4.6' and its new context window.

Activity

All 870 unit tests passed, with 14 skipped and 0 failures, ensuring stability.
Alias resolution tests were updated and confirmed to cover all newly added models.
Backward compatibility for version-specific aliases was thoroughly verified.
A check confirmed no duplicate aliases exist across the 76 total aliases.
Model IDs were successfully verified against the live OpenRouter API.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

Keep version 9.9.0 from our feature branch (upstream synced to 9.8.2). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

gemini-code-assist

Code Review

This pull request adds 9 new models, removes 2 obsolete models, and updates various aliases and tests accordingly. The changes are extensive and mostly look good. However, I've found a few issues that need attention. In pyproject.toml, the required Python version is bumped to >=3.10 without justification, which is a breaking change. Additionally, in conf/openrouter_models.json, there are a couple of configuration issues: one new model has a max_output_tokens value that leaves no room for an input prompt, and another has a contradictory setting for temperature support. Please see my detailed comments for suggestions.

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f428374514

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

- gpt-5.3-codex: supports_temperature true→false (contradicts fixed constraint) - minimax-m2.5: max_output_tokens 196608→32768 (practical default; API allows up to 196K) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

timeleft-- · 2026-03-06T14:30:55Z

Addressed Review Feedback

Fixed in 830f732:

Temperature contradiction (gpt-5.3-codex) ✅

Changed supports_temperature from true to false — contradicted temperature_constraint: "fixed".

Note: The same contradiction exists in pre-existing models (gpt-5-mini, gpt-5-nano, gpt-5.2, gpt-5.1-codex, gpt-5.1-codex-mini). Those are out of scope for this PR but could be a good follow-up.

MiniMax M2.5 max_output_tokens ✅

Changed from 196608 to 32768. OpenRouter API reports max_completion_tokens=196608 (matching context_window), but a practical default of 32K leaves room for input prompts. Added note in description field.

requires-python >=3.10 — Intentional, not accidental

This change is necessary: mcp>=1.0.0 (a core dependency) requires Python >=3.10. Running uv lock with >=3.9 fails:

No solution found when resolving dependencies for split (python_full_version == '3.9.*'):
Because mcp>=1.0.0 depends on Python>=3.10...

The project has never actually worked on 3.9 due to this transitive requirement. The requires-python metadata was simply incorrect.

gpt5pro alias cross-provider drift — Acknowledged

Valid observation. conf/openai_models.json still maps gpt5pro → gpt-5.2-pro, and OpenAI native takes priority over OpenRouter. This PR focuses on the OpenRouter catalog; updating the OpenAI native config (adding gpt-5.4-pro) would be a separate PR.

10 files were failing the black --check CI step before this PR. None are files modified by this PR — fixing them here to unblock CI. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- Add openai/gpt-5.4 with aliases gpt5, gpt5.4, gpt-5.4 - Move gpt5 alias from openai/gpt-5 to openai/gpt-5.4 (newest base) - Keep gpt-5.0/gpt5.0 on old openai/gpt-5 for backward compat - Update tests and docs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- codex → openai/gpt-5.3-codex (was gpt-5-codex) - Old gpt-5-codex keeps codex-5.0 for backward compat Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

…hiveInnovations#411) Cherry-picked model additions from BeehiveInnovations#411. New models: Claude 4.6 (Opus/Sonnet), Gemini 3.1 Pro, GPT-5.4/5.4-Pro, GPT-5.3-Codex, Devstral, DeepSeek V3.2, Qwen 3.5, MiniMax M2.5. Updated generic aliases (opus→4.6, sonnet→4.6, pro→3.1, gpt5→5.4, codex→5.3) with version-specific aliases for backward compatibility. Fixed no-API-keys test to account for ADC fallback from PR BeehiveInnovations#306.

illera88 · 2026-03-13T23:11:08Z

ship it! 💯

JCMais · 2026-03-30T11:57:53Z

@guidedways Can we get this in pls? Looking forward to using gpt 5.4 with consensus / codereview

…im to 5 core tools Merged community PR BeehiveInnovations#411 adding GPT-5.4, Gemini 3.1 Pro, Claude Opus 4.6, Grok 4.1, and 5 other frontier models to the OpenRouter catalog. Deleted 13 tools, keeping only the 5 we use: - consensus (multi-model debate) - chat (direct model access) - debug (root cause analysis) - thinkdeep (extended reasoning) - codereview (structured code review) Removed: precommit, planner, challenge, apilookup, analyze, refactor, testgen, secaudit, docgen, tracer, version, listmodels, clink. Each replaced by existing tools in our stack (Context7, deploy hooks, EnterPlanMode, CLAUDE.md directives, PreToolUse hook). 9,757 lines removed. Clean, focused fork. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

timeleft-- · 2026-04-14T21:24:24Z

Closing — all changes from this PR are already merged into the MachineWisdomAI fork and superseded by #430 (April 2026 model refresh).

merge: Resolve config.py version conflict with upstream main

8edb2f7

Keep version 9.9.0 from our feature branch (upstream synced to 9.8.2). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

gemini-code-assist bot reviewed Mar 6, 2026

View reviewed changes

Comment thread conf/openrouter_models.json Outdated

Comment thread conf/openrouter_models.json Outdated

Comment thread pyproject.toml

chatgpt-codex-connector bot reviewed Mar 6, 2026

View reviewed changes

Comment thread pyproject.toml

Comment thread conf/openrouter_models.json

fix: Address review feedback on model configs

830f732

- gpt-5.3-codex: supports_temperature true→false (contradicts fixed constraint) - minimax-m2.5: max_output_tokens 196608→32768 (practical default; API allows up to 196K) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

timeleft-- and others added 3 commits March 6, 2026 09:32

style: Run black on pre-existing unformatted files

e96c33c

10 files were failing the black --check CI step before this PR. None are files modified by this PR — fixing them here to unblock CI. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

feat: Migrate codex alias to latest (gpt-5.3-codex)

35da1ca

- codex → openai/gpt-5.3-codex (was gpt-5-codex) - Old gpt-5-codex keeps codex-5.0 for backward compat Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

timeleft-- closed this Apr 14, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: Add 9 new models, remove 2 obsolete, update aliases#411

feat: Add 9 new models, remove 2 obsolete, update aliases#411
timeleft-- wants to merge 6 commits intoBeehiveInnovations:mainfrom
MachineWisdomAI:feat/update-openrouter-models-2026-03

timeleft-- commented Mar 6, 2026

Uh oh!

gemini-code-assist bot commented Mar 6, 2026

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Uh oh!

Uh oh!

Uh oh!

timeleft-- commented Mar 6, 2026

Uh oh!

illera88 commented Mar 13, 2026

Uh oh!

JCMais commented Mar 30, 2026

Uh oh!

timeleft-- commented Apr 14, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

timeleft-- commented Mar 6, 2026

Summary

Details

New models

Alias migration (backward compatible)

Test plan

Uh oh!

gemini-code-assist bot commented Mar 6, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

timeleft-- commented Mar 6, 2026

Addressed Review Feedback

Temperature contradiction (gpt-5.3-codex) ✅

MiniMax M2.5 max_output_tokens ✅

requires-python >=3.10 — Intentional, not accidental

gpt5pro alias cross-provider drift — Acknowledged

Uh oh!

illera88 commented Mar 13, 2026

Uh oh!

JCMais commented Mar 30, 2026

Uh oh!

timeleft-- commented Apr 14, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants