**TL;DR:** The provider does not recognize the model name. It is almost always a typo in the model string, or a proxy or wrapper in front of the provider that only allows certain model names. Copy the string from the docs, do not type it from memory. If you use init_chat_model, try the explicit provider:model form ("openai:gpt-5.5") so there is no ambiguity about which provider should resolve the name.

## The fix

1. Spell-check the model identifier character by character. `gpt-5.5` vs `gpt-5-5`, `claude-sonnet-4-6` vs a made-up variant. Copy the string from the docs, do not type it from memory.

2. If you use `init_chat_model`, try the explicit `provider:model` form (`"openai:gpt-5.5"`) so there is no ambiguity about which provider should resolve the name.

3. Behind a proxy, gateway, or model wrapper? Confirm the permitted model names on that layer. Proxies often restrict or rename models, and the error then comes from the proxy, not the provider.

4. Brand-new model names need a recent provider package. An old langchain-openai may not know a model released last month.

## When to use this

- You are seeing this exact error message; match the block above, not just part of it.
- The failing call matches the scenario in the title: MODEL_NOT_FOUND.
- You want the fastest verified fix before digging through logs.

## When not to use this

- Your error text differs from the block above; close cousins often have different causes.
- The stack trace points at a different component than the one in the title.
- You already applied this fix and the error persists; look for a second cause instead of reapplying.

## Compatibility

- Mentions 5.5 in the original; behavior may differ on much older or newer releases.