What is multi-update-models?

A procedure for checking the newest API model names from Anthropic, OpenAI, and Google and updating a project's model settings.

Use it to inspect the project's current model aliases and defaults, research each provider's latest models, and update the relevant configuration files.

Model names and availability change over time, so configuration files can become outdated or point to models that are no longer preferred or available.

multi-update-models — Command by religa

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add commands/religa/multi_mcp/multi-update-models

Clone the repo

git clone --depth 1 https://github.com/religa/multi_mcp

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 910 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00000	$0.00910
Opus 5	$0.00000	$0.00455
Sonnet 5	$0.00000	$0.00182
Haiku 4.5	$0.00000	$0.00091

Measured 2d ago against content hash 48fc0542f7ea, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

multi-update-models scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/multi-update-models.md · 81 lines

How it starts

The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Update Model Configuration

Research the latest available models from Anthropic, OpenAI, and Google, then update the project's model configuration accordingly.

Step 1: Read Current Config

Read the current model configuration:

multi_mcp/config/config.yaml - model definitions, aliases, constraints
multi_mcp/settings.py - default model and default_model_list fallback

Note which models and aliases are already present.

Step 2: Research Latest Models

Use web search to find the latest available API models from each provider. Search the general web (e.g., "Anthropic latest Claude models API 2026", "OpenAI latest models API", "Google Gemini latest models API") in addition to checking provider docs. The doc URLs below are starting points but may become stale - always verify via general web search as well.

Anthropic: Start with https://docs.anthropic.com/en/docs/about-claude/models - also search the web for latest Claude model announcements and API model IDs
OpenAI: Start with https://platform.openai.com/docs/models - also search the web for latest GPT and o-series model announcements and API model IDs
Google: Start with https://ai.google.dev/gemini-api/docs/models - also search the web for latest Gemini model announcements and API model IDs

For each provider, determine:

New models not yet in config.yaml
Models that have been deprecated or shut down
Changes to existing model capabilities (context window, output tokens, etc.)

Step 3: Update Config

Edit multi_mcp/config/config.yaml:

Add new models with appropriate litellm_model strings (prefixed with provider: openai/, anthropic/, gemini/)
Remove models that have been shut down (not just deprecated - only remove if actually unavailable)
Move aliases (like mini, nano, gpt, sonnet, gemini) to point to the latest generation
Mark superseded models as legacy in their notes field
Set provider_web_search: true for models that support it
Set temperature constraints (e.g., temperature: 1.0 for GPT-5 family)
Keep the existing section structure (OpenAI, Azure, Anthropic, Google, CLI, Bedrock, OpenRouter)

Read the full file on GitHub · 81 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2d ago First seen · 81 lines · 0 tokens per session scan A 48fc0542f7ea

Subscribe to this mod's changes

multi-update-models is a command published in the GitHub repository religa/multi_mcp (35 stars, last pushed 19d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 910 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Backend and APIs Data and AI LLM Research claude claude-skills developer-tools llm-orchestration mcp

prep

Purpose: Button up everything at the end of a session — or before a /compact. Memories, plans, git — all tidy. Works for both closing out a chat and preparing for compaction.

AIOSAI/AIPass · 0 tokens

compass

Purpose: Capture the decision just made into compass (the rated decision engine) with the user's rating and note. The user fires this when they notice a decision worth recording — they supply the judgement, you supply the decision text from the conversation. This is the human-triggered answer to the "noticing"…