When a Bedrock-style proxy returns 501 for count_tokens, the client no longer falls back to a Haiku max_tokens:1 probe. It instead counts tokens with a one-token request using the session's own model, unless ANTHROPIC_SMALL_FAST_MODEL or ANTHROPIC_DEFAULT_HAIKU_MODEL is set.
Names in the bundleANTHROPIC_SMALL_FAST_MODELANTHROPIC_DEFAULT_HAIKU_MODEL
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Confirmed since
Anthropic's documentation has since written up ANTHROPIC_SMALL_FAST_MODEL, on How Claude Code uses prompt caching.
The variable also covers a background model you set with the deprecated `ANTHROPIC_SMALL_FAST_MODEL` variable, when that model differs from your main model.prompt-caching see the edit
Confirmed since
Anthropic's documentation has since written up ANTHROPIC_DEFAULT_HAIKU_MODEL, on Connect Claude Code to an LLM gateway.
* **Model versions**: pin `ANTHROPIC_DEFAULT_OPUS_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, and `ANTHROPIC_DEFAULT_HAIKU_MODEL` as in [Pin model versions](/docs/en/google-vertex-ai#5-pin-model-versions). Setting `ANTHROPIC_DEFAULT_HAIKU_MO…llm-gateway-connect see the edit
One source agreesOne thing we can check says the same as this entry.
Anthropic's documentation agrees
Anthropic's documentation has since written up ANTHROPIC_DEFAULT_HAIKU_MODEL, on Connect Claude Code to an LLM gateway.